New York City Independent Budget Office

The Fiscal Impact of the
New Federal Welfare Law
on New York City



Notes



Table of Contents


Summary


The enactment of Public Law 104-193, the Personal Responsibility and Work Opportunity Reconciliation Act, on August 22, 1996, signals the end of federal welfare assistance as an entitlement to individuals. Instead of providing federal aid to individuals based on specific eligibility standards, a new block grant program is established to provide states with funds to tailor income maintenance programs to the states' particular public assistance needs. In addition to many other new requirements, the Act denies federal assistance to specific categories of individuals, such as persons receiving assistance for a total of five years, and to certain legal aliens.

States are required to develop plans during 1997 to implement the new law. While a number of restrictions are placed on the use of block grant funds, the Act provides a great deal of discretion for states to determine the precise makeup of its public assistance programs. Like other states, New York State must make many decisions, through both legislative and executive branch processes, to alter its present approach to welfare assistance. Among other things, these decisions involve the allocation of funds to localities, implementation of new work requirements, and treatment of qualified immigrants.

Until these decisions are made by the State, it is difficult to assess the impact of the Act on New York City's budget. Nevertheless, the Independent Budget Office (IBO) has prepared this report to provide an initial projection of costs that could be incurred by the City in the wake of the Act. This report does not attempt to predict future policy decisions by the State government, but it does highlight a number of critical decisions (e.g., funds allocation, work requirements) yet to be made and their potential fiscal impacts on the City.

Three scenarios are used throughout the report to illustrate potential costs to the City. Such cost estimates are very sensitive to assumptions concerning trends in New York City economic performance, demographic profile, immigration, and implementation dates established by the State. One scenario assumes continued moderate economic performance so that any change in the number of welfare cases can be directly attributed to the Act and New York's response; a second assumes weaker, albeit positive, growth resulting in more public assistance cases than otherwise would result; and a third assumes stronger economic growth contributing to fewer cases.

If the set of assumptions used to prepare this report is realized, P.L. 104-193 would result in substantial additional costs to the City through 2002 when the Act faces reauthorization. Although the new law would have little impact on the City's budget through 1998, costs begin to rise dramatically in 1999. Under the most optimistic scenario considered by IBO, the Act would result in a total of nearly $200 million of additional City expenditures from 1997 to 2002. At the other end of the range of scenarios, costs could run as high as $4.3 billion over the six year period. Given uncertainties regarding future local economic performance and implementation details of the State, it is impossible to predict exact costs. The sheer magnitude of these potential costs, however, makes it clear that the City could be confronted with tremendous budgetary challenges to meet the basic public assistance needs of its citizens in the years ahead.

Program Changes Resulting from the Act

Provisions of the Act potentially having the greatest bearing on the New York City budget are as follows:

Assumptions Regarding State's Response

The TANF block grant is the cornerstone of the federal welfare reform effort. This approach to providing public assistance is fundamentally different from entitling individuals to assistance at the federal level. Under the AFDC program, states had relatively little discretion to decide eligibility, work requirements, and the length of time benefits would be provided. Block grants, on the other hand, are designed to provide states with flexibility by setting broad eligibility criteria and leaving implementation details to the state. Accordingly, states now must make many more decisions concerning the administration of programs to assist low-income persons than in the past.

A number of the decisions to be made by the New York State government will inevitably have a profound impact on the New York City budget. The most significant of these decisions involves the following:

It should be noted that certain social costs could be incurred if the State does not extend Home Relief assistance to those individuals dropped from the TANF rolls. The prospect of a substantially weakened social safety net for poor City residents increases the likelihood for higher costs in the future for, among other things, homeless shelters, foster care, criminal justice, and medical care. These costs have not been factored into this analysis because it is assumed that the State and City, while requiring more work for welfare recipients, will continue to maintain public assistance near its current levels.

Estimated Direct Costs to New York City

Caseload Projections

Any estimate of the impact of the Act on New York City will critically depend not only on decisions made by the State but also on future economic performance. In order to provide a range of possible costs to the City, this report contains three scenarios: one assumes continued moderate economic performance so that any change in the number of welfare cases can be directly attributed to the Act and New York's response (the moderate caseload scenario); a second assumes weaker, albeit positive, growth resulting in more public assistance cases than otherwise would result (the higher caseload scenario); and a third assumes stronger economic growth contributing to fewer cases (the lower caseload scenario). Because the same rate of economic growth is assumed as in the City's adopted budget financial plan, the moderate caseload scenario is best suited for cost and caseload comparisons with financial plan projections.

As illustrated in Figure S-1, beginning in 1998 about 60,000 more New York City residents are expected to receive public assistance each year through TANF and Home Relief than would have received aid through AFDC and Home Relief under prior law. Through 2001, this net gain in cases results from a substantial increase in Home Relief cases (TANF remains at about the same level as AFDC) attributable to an influx of legal aliens previously eligible for various forms of federal assistance. Beginning in 2002, however, the mix of recipients shifts in response to a substantial decrease in TANF cases and a surge in Home Relief cases as TANF recipients begin to reach the five year limit of assistance. In other words, assuming no changes in eligibility for Home Relief, eliminating low-income individuals from TANF assistance simply shifts cases and costs from the federal government to the City and State.

Work Requirements and Child Care Cost Implications

Two distinct work requirements are established in the Act for adult recipients of TANF. First, states are required to ensure that 25 percent of recipients gain employment in 1997, rising by 5 percentage points each year through 2002. States unable to meet these quotas face significant fiscal penalties. Second, as noted above, work is mandated for all recipients within two years of first receiving assistance. There is no penalty, however, on states for non-compliance with this provision.

These work requirements are expected to have substantial cost implications for the City. Based on an analysis of workfare programs in New York and elsewhere, IBO estimates that it would cost $2,250 per worker in 1997 and rising each year thereafter to pay for administrative overhead, transportation, supervision, and other factors associated with placement of these new workers. Accordingly, if the two year work rule is enforced, many new workers will need to be placed more quickly, resulting in substantially higher costs to the City than if the rule is ignored.

These ambitious work requirements will require a major commitment on the part of the federal, State, and City governments to ensure that adequate child care is provided for newly working families. Based on an analysis of age distribution of children of AFDC recipients, projected TANF caseloads, and the work requirements described above, IBO estimates that between 33,000 and 84,000 full-time child care slots will be needed by 2002, depending on whether or not the two year rule is enforced. At an average cost of $5,000 per slot in 1997 and rising each year thereafter, these child care costs will likely run into the hundreds of millions of dollars by the year 2002.

Total Costs Under the Moderate Caseload Scenario

Most of the cost to the City's budget results from two factors: the shift of welfare recipients from federally subsidized assistance to programs entirely funded by the State and City, and the costs of providing jobs and child care to public assistance recipients. Some of these costs are partially offset, however, by the size of the TANF grant to New York State and the resulting City share. In fact, at least in the near term, the TANF grant will exceed the amount that would have been available to the State under the AFDC program. The difference between these two amounts, sometimes referred to as a "surplus" or "windfall", represents an additional amount that could be made available to the City (depending on State decisions) to help defray the costs of implementing the Act.

Figure S-4 illustrates estimated costs to the City under the moderate caseload scenario resulting from the Act, assuming the two year work rule is not enforced. The hatched and shaded bars represent total City costs while the white bars show the City's share of federal TANF block grant funds remaining after assistance is provided. The difference between these bars represents the net additional cost to the New York City budget and is shown in Figure S-6. Although net costs are minimal over the first two years of the Act's implementation, cost begin to rise dramatically in 1999 and ultimately total $628 million over the 1997 to 2002 period.

As shown in Figure S-5, these costs become even larger if the two year work rule is enforced. Net costs over the six year period total nearly $3.8 billion. Most of this increased cost is attributable to the higher costs of funding worker placement and child care programs. As discussed earlier, it is unclear whether the State would choose to enforce the two year work rule given both the absence of any fiscal penalty for non-compliance and the expected additional costs the State and its localities.

A comparison of net costs to the City of these two variations of the moderate caseload scenario is shown below as Figure S-6. This figure clearly illustrates the substantially higher costs that would be incurred if the two year work rule is enforced. It should be emphasized that these are net additional costs to the City, over and above amounts already included in the adopted budget financial plan.

Total Net Costs Under the Higher and Lower Caseload Scenarios

As shown in Figure S-7, costs to the City rise dramatically under the higher caseload scenario particularly if the two year work rule is enforced. Even without enforcement, additional costs to the City would total $563 million by 2002. The higher caseload scenario would likely be realized (in the near term) only in the event of a mild economic downturn. On the other hand, the lower caseload scenario would result in substantially lower additional costs to the City. In fact, costs would be minimal if the two year rule is not enforced. Such a scenario would result from more robust local economic growth coupled with an increase in welfare recipients securing unsubsidized private sector employment, possibly in response to the Act's welfare to work provisions.

Implications for State-City Share

It is important to note that total State costs for welfare programs in New York City will vary much less and will generally be significantly lower than City-funded costs. This is a result of two factors. First, the City will bear the burden for almost all reform related workfare and child care costs, less any costs covered by federal TANF funds available after paying for income maintenance. Second, the amount of TANF funds available for workfare and child care in New York City will be disproportionately low if the State allocates only 60 percent of the block grant to City programs. These disparities between the City and State shares of welfare costs would be substantially reduced, and the City's fiscal risk under welfare reform would be significantly eased, if New York City were to receive a TANF allocation in line with its 68 percent share of total AFDC caseloads.

Welfare Reform: A Longer View

The new welfare law's most significant fiscal impacts are likely to occur in the years 2002 and beyond. Important implications for future City budgets include the following:

Conclusion

The Act's impact on the New York City budget will ultimately depend on the decisions of State and City officials, the performance of the local economy, and the behavior of thousands of individuals in response to the new system. These uncertainties coupled with the newness of the program make estimating the fiscal impact of the new law particularly difficult.

Assuming that the City receives 60 percent of the State's TANF block grant, IBO projects that the Act will have little impact on the City's budget over the next two years. Beyond 1998, costs could rise significantly depending on caseload. The timing of these costs depends on whether or not the State chooses to enforce the two year work rule or to phase in federal work requirements more gradually. Mandating welfare recipients to work has the effect of significantly increasing the City's cost of administering work programs and providing child care.

Under the moderate caseload scenario, IBO projects that the City's cost of providing welfare would increase by $33 million in 1999 growing to $269 million by 2002. If the two year work rule is enforced, these additional costs would total $723 million in 1999 and roughly $1 billion annually through 2002.

Under the higher caseload scenario, additional costs would increase by $38 million in 1999 growing to $563 million by 2002. If the two year work rule is enforced, estimated additional costs would total $733 million in 1999 and more than $1.2 billion by 2002.

In contrast, lower caseloads could result from some combination of more robust economic growth and the law's potential success in moving welfare recipients off the rolls and into private sector employment. Under a lower caseload scenario, the City could expect to incur additional costs of less than $100 million annually. If the two year work rule is enforced, additional costs would range from $700 million to $942 million over the 1999-2002 period.

Further, it is important to emphasize that the State's decision on how the TANF block grant will be shared is a critical factor in determining the cost to New York City. The City currently accounts for 68 percent of all AFDC recipients. If New York City were to receive 68 percent of the TANF block grant as opposed to the 60 percent assumed in the analysis then the City would receive an additional $190 million each year thereby dramatically reducing expected costs.

Given the fact that much of the cost discussed in this report results from worker placement and related child care costs, the price tag of the new law could be substantially lowered if the local economy is able to grow at a rate sufficient to create and sustain employment for the tens of thousands of welfare recipients in need of work.

Although future economic performance is unknown, it is clear that the new welfare law's most significant fiscal impacts will be felt in the year 2002 and beyond, as a greater share of public assistance is borne by the City and State. Given existing and projected budgetary difficulties, such increased costs will make it ever more difficult for the City to meet the basic needs of its poorest citizens.

Introduction

The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 (Public Law 104-193) recently signed into law represents the most radical change in the welfare system in the last sixty years. These changes in welfare will affect the lives of all New Yorkers, especially those receiving public assistance and those living in the City's poorest neighborhoods.

Several sections of the new law will have a significant impact on New York City's budget. Among these are provisions to: replace Aid to Families with Dependent Children (AFDC) with a new Temporary Assistance for Needy Families (TANF) state block grant; impose a five year lifetime limit on families receiving cash benefits; mandate increasingly ambitious work quotas for adult family heads; provide increased child care funding; restrict eligibility of children for Supplemental Security Income; and limit benefits for legal aliens. A summary of these changes is shown in Figure 1.

This report examines the impact of the Act on the New York City budget1. The first section provides an overview of current welfare programs, the City's own efforts to reduce welfare spending, and the States' constitutional requirement to aid the needy. The next section provides estimates of the direct fiscal impact of the new welfare law through the year 2002, first for a moderate welfare caseload scenario and then for alternative higher and lower caseload scenarios. The third section considers the impact of the Act in 2002 and beyond. A concluding section briefly summarizes major findings.

The Local Social Service Context of Federal Welfare Reform

Before analyzing the budgetary impact of the new law on the City, it is important to review both the scope of the City's current social service spending and the policy and legal context in which federal reform occurs. New York City has long been known for having one of the most comprehensive social welfare systems in the United States, and the various forms of government-funded welfare and health services play a significant role in the City's budget and economy. The extent of this role is discussed in the next section.

This overview is followed by a review of the City's own recent welfare reform efforts, which already have had a significant impact on caseloads and City spending. The section concludes with a brief discussion of the State constitution's mandate to provide support for the needy, which is likely to shape New York's response to the Act.

Current Welfare Programs and Spending

The scope of current welfare programs and spending in New York City can be seen through a snapshot of the social services delivery system in the most recent month for which complete data is available, May 1996 (New York State Department of Social Services, 1996a).

Aid to Families with Dependent Children (AFDC) provides federally mandated entitlement cash assistance to low-income households with children under the age of 18. In May 1996, a total of 810,500 City residents received almost $147 million in AFDC payments. Half of AFDC funding comes from the federal government, and in New York State the non-federal portion is split evenly between the State and localities. In its 1997 adopted budget, the City appropriated nearly $1.7 billion for AFDC, of which $425 million (25 percent) is funded by the City itself.

Home Relief is a New York State cash assistance program for low-income households that do not qualify for AFDC, usually because they contain no children under 18 years of age. In May 1996, a total of 214,500 City residents received over $52 million in Home Relief payments. Funding of the Home Relief program is split evenly between the State and City. In its 1997 adopted budget, the City appropriated $524 million for Home Relief, half of which represents City funds.

Supplemental Security Income (SSI) provides federally mandated cash assistance to aged, blind, or disabled low-income persons. Approximately 372,000 persons in New York City received SSI payments totaling $156 million in May 1996. Most SSI funds are federal, though there is a small State supplement. In 1995, City residents received a total of $1.7 billion in SSI payments.

Food Stamps provide federally mandated benefits to low-income households. Recipients are allocated coupons to purchase food at grocery stores and other outlets. In May 1996, a total of 1,345,400 City residents received food stamps worth $115 million. The food stamp program is funded entirely by the federal government. In 1995, food stamp payments in the City totaled $1.4 billion.

Medicaid is the main government program to fund medical care for those unable to afford private care. Medicaid eligibility is generally granted to those receiving cash assistance through the AFDC, Home Relief or SSI programs. In May 1996, a total of 1,944,000 clients received nearly $1.4 billion in Medicaid benefits. Unlike the AFDC or Home Relief programs where payments to recipients are made by the City, Medicaid payments are made by the State, although non-federal costs are shared by both the City and State. The City budgets only its portion of Medicaid spending, and the percent of this portion varies greatly, depending on the eligibility category of the recipient and the type of medical service provided. In its 1997 adopted budget, the City appropriated $2.6 billion of its own funds for Medicaid.

Figure 2 summarizes social service spending by program in New York City for calendar year 1995 for all levels of government.

City Welfare Reform Efforts

Faced with growing budget deficits, City officials have implemented a number of measures in recent years designed to slow and possibly reverse the growth in the costs of entitlement programs. The most significant of these has been the New York City Work, Accountability and You program (NYC WAY). Begun in January 1995, NYC WAY incorporates enhanced efforts to detect and prevent welfare fraud as well as comprehensive work programs for able-bodied adult recipients of public assistance. Fraud detection measures have included an electronic finger imaging system for welfare clients and additional eligibility verification procedures including visits to the homes of new applicants and current recipients.

The City's Human Resources Administration (HRA) began applying these initiatives to Home Relief applicants in January 1995 and to current recipients the following July. The program has been the primary factor contributing to a significant caseload decline, from 295,715 persons in January 1995 to 209,859 in June 19962. In recent months, HRA has begun to expand application of the NYC WAY provisions to the AFDC population. As a result of these local welfare reform efforts, the City's Office of Management and Budget (OMB) has projected continued caseload declines in City fiscal year 1997 of about 3,000 persons per month for Home Relief and 5,600 persons per month for AFDC. So far in calendar year 1996, the pace of caseload reductions in the City has exceeded OMB's projections. The effectiveness of NYC WAY and other local efforts in the future will influence welfare caseloads and thus the budgetary impact of federal reform.

The New York State Constitution

In analyzing the future impact of federal welfare reform efforts on New York City, it is also important to consider the State constitutional context. Article XVII, Section 1 of the New York State constitution requires the State and City to provide for "the aid, care and support of the needy." While this provision offers some protection to the integrity of the safety net in the face of budget pressures, it may have the effect of limiting the options available to State and City officials in implementing the requirements of the new federal statute.

The case law interpreting Article XVII is limited and inconclusive. However, the case law that does exist suggests that implementation of two provisions of the new law may be especially vulnerable to challenge: the denial of benefits to legal aliens and the five year time limit on cash assistance to families. In both cases it is possible that the courts will require the State and City to maintain aid for the affected groups in spite of the loss of federal funding, thus increasing the costs of the Act on the State and City budgets.

It is within these budget, policy, and legal contexts that the City will begin to implement the provisions of the new federal law. The Direct Impact of the Act on the New York City Budget:

Annual Estimates Through 2002

This section examines the direct impact or total net fiscal cost of the Act on the New York City budget. The analysis covers the period through 2002, the year that the Act will be up for renewal. A discussion of post-2002 impacts is included in a later section that suggests that long term costs could be particularly high.

Any projection of the impact of the Act will depend critically on a number of assumptions about the future, not the least of which is the forecast of public assistance caseloads for the coming period. The first part of this section presents three alternative projections through 2002 of potential AFDC caseloads in New York City absent federal reform. The next part details various programmatic and funding changes affecting the City's costs for an intermediate scenario of caseload projections. The final part of the section presents comparable calculations for alternative high and low projections.

As the subsequent discussion makes clear, regardless of particular caseload projections, a number of other assumptions must be made to estimate the budgetary impact of the new law. Two of the most significant concern the share of the federal block grant the State will allocate to the City and the division of non-federal spending between the City and State. These are discussed below.

The IBO's analysis assumes that the courts will interpret the State constitution to require the State to aid those who are needy yet no longer qualify for the TANF program, and that the State will do so by providing aid to them through the Home Relief program at existing benefit levels.3 Other State legislation (including that needed to implement the TANF program), administrative regulations, and judicial rulings will also influence the size of the budgetary impact.4

In view of all the uncertainties affecting the calculations of the direct impact of the new law, IBO has calculated a range of plausible total net costs to the New York City budget of the Act, rather than a specific prediction.

Public Assistance Caseload Projections

The first step in analyzing the fiscal impact of the Act is to estimate caseloads through 2002. In order to provide an analytic framework for cost comparisons, three estimates of AFDC caseloads in the absence of welfare reform were developed based on different economic scenarios. The purpose was to provide projections that could be used to analyze the fiscal impact of the Act; they are not intended as specific predictions of future caseloads. This section briefly describes the three caseload estimates and the assumptions behind them.

Without policy or law changes, the rate of job creation or loss, (which in turn is closely related to the rate of economic growth), is the primary factor contributing to changes in public assistance caseloads. Slow or negative economic growth accompanied by slower increases or absolute declines in the number of workers employed leads to higher welfare caseloads. Conversely, robust growth and gains in employment eventually yield lower caseloads. In New York City, such relationships have held up over time.5

The Mayor's Executive Budget for City fiscal year 1997 projects moderate economic growth over the coming years, averaging 2.24 percent annually for the calendar years 1996 through 2000. This moderate growth projection provides the foundation for IBO's intermediate caseload projection. Accounting for another year of reductions attributable to the local NYC WAY initiative, the budget projects that AFDC caseloads will fall to 744,000 by June 1997 and then remain constant for the remaining years of the financial plan. For purposes of the moderate growth caseload estimate, IBO extends this constant caseload forecast through 2002, as indicated in Figure 3.

The higher and lower caseload scenarios are the same as the moderate projection for the first two years, but in 1999, they begin to diverge. The higher caseload scenario projects that caseloads will gradually begin to increase, rising by 0.8 percent to 750,000 in 1999, followed by significant increases of 3.3 percent to 775,000 in 2000, and 3.9 percent to 805,000 in 2001. The rate of increase then moderates to 1.9 percent in 2002 with the caseload reaching 820,000. For the period 1998 through 2002, the caseload is projected to grow by 76,000 (10.2 percent) under this scenario. This pattern would be consistent with a mild recession in the City's economy similar to the downturn that occurred in 1981 and 1982.

In contrast to the caseload increases forecast under slower economic growth, faster economic growth contributes to a decline in caseloads. From 1998 to 1999, caseload drops by 1.2 percent to 735,000, by 2.0 percent to 720,000 in 2000, by 2.8 percent in 2001, and by 4.3 percent in 2002 when the caseload falls to 670,000. For the period 1998 through 2002, the caseload under this scenario would decrease by a total of 74,000 (9.9 percent). These declines would be of smaller magnitude than those observed in New York City's 1983 to 1988 expansion.

With these alternate projections of AFDC caseloads, it is possible to calculate a range of direct impacts of the Act on the New York City budget. The block grant method of federal funding, the imposition of cumulative time limits, work requirements and other features of the reform package will all affect the total net cost to the City of federal reform. The following two sections of this report will examine welfare reform's direct budgetary impact, first for the moderate caseload scenario and then for the higher and lower alternative scenarios.

Direct Fiscal Impact Under The Moderate Caseload Scenario

In deriving its estimates of the total net fiscal cost of the Act, both for the moderate caseload scenario presented here and the alternative scenarios presented in the next section, IBO has assumed that the current policies determining the allocation between the City and the State of non-federal costs of Home Relief and other programs will continue. Much of the budgetary impact of the new law comes from a shifting of cases to and from social service programs where the mix of federal, State and City funding differs.

The primary feature of federal welfare reform is the replacement of the AFDC program with the new TANF program. The TANF program incorporates into one block grant federal funds formerly budgeted through separate appropriations for the AFDC grant program, AFDC administrative costs, Emergency Assistance to Families (EAF) and the Job Opportunities and Basic Skills Program (JOBS). Block grants will be distributed to the states based on each state's levels of spending for these programs in fiscal years 1992 through 1995.

Significantly, with respect to the determination of the size of the block grants, the law makes no allowance for inflation. While the legal entitlement for a qualifying family to participate in these programs is eliminated, states have greater flexibility to mold the TANF program to fit their own needs. It is important to note, however, that the Act includes a maintenance of effort requirement that limits the ability of states to use block grant funds for purposes other than implementing the provisions of the Act.6

In addition to estimating the fiscal impact on the City of changes in the TANF caseload and the size of the TANF block grant, this section will also examine the impact of the Act's provisions concerning work requirements, associated child care costs, and other stipulations affecting legal aliens.

TANF Caseload Changes

Before examining the size of the block grant in relation to New York's funding needs, it is important to analyze how the programmatic changes of the Act will affect the caseload of the basic TANF program. There are three changes that need to be considered. The provision with the largest effect on the caseload is the five year cumulative limit for TANF recipients. In addition, changes in the eligibility of children for the largely federally funded SSI program will add to the TANF caseload. The Act also restricts eligibility for newly arriving immigrants, thereby reducing the TANF caseload.

The Five Year Time Limit of TANF

One of the most significant provisions of the new welfare law is the five year time limit. States are required to drop recipients from their TANF rolls once they have received a cumulative total of five years of aid since enactment of the Act. States are allowed to exempt up to 20 percent of their caseload from this time limit for reasons of hardship, though, conversely, they are also authorized to impose a shorter time limit.7

The time limit provision could result in substantial additional costs for the City and State in 2002 and beyond. Using the data from the Urban Institute's on the distribution of AFDC recipients' length of stay on welfare (Pavetti, 1995), IBO estimates that in 2002, as the first cohort of recipients reaches its limit, 31 percent of existing TANF clients will have received benefits for at least five years.

Based on this estimate, the cost to New York City of moving clients who have reached the TANF time limit onto the more costly (for the City) State Home Relief program could be as much as $123 million per year under the moderate caseload scenario.8 This cost would result if the State does not exercise the option of exempting a portion of the caseload from this limit. However, if the State does exercise its exemption to the greatest extent possible, only 11 percent of all TANF recipients or approximately one third of those who would reach the five year limit in 2002 would be denied TANF benefits and shifted to Home Relief.9

For the purposes of calculating the fiscal impact of welfare reform through the year 2002, it is assumed that the State will fully exercise the 20 percent exemption allowance., Accordingly, the cost to the City in 2002 of absorbing 11 percent of TANF recipients into Home Relief is projected to be $45 million. In subsequent years, however, greater numbers of TANF recipients will have received assistance for a cumulative total of at least five years. The growing fiscal impact in the post-2002 period of the cumulative time limit for AFDC is discussed below.

Supplemental Security Income for Children

The Act decreases the number of children eligible to receive assistance under the SSI program. This program provides assistance to blind and disabled children and adults who meet certain income qualifications. Under prior law, children qualified for SSI in one of two ways: they may have had a debilitating condition such as cystic fibrosis or asthma, or they may have been determined to have impaired functioning below that which is normal for children of their age as determined by a state-conducted Individual Functional Assessment. The new welfare law eliminates the Individual Functional Assessment and the consideration of maladaptive behavior as a disability. Only those children who have a medically determinable physical or mental impairment which results in marked and severe functional limitations expected to result in death, or expected to last for at least 12 months, will now be considered disabled.

The new law requires the federal Social Security Administration (SSA) to review cases of children currently receiving SSI to determine whether or not recipients meet the new guidelines. The earliest date children currently receiving SSI would lose benefits is July 1, 1997, and SSA must notify all potentially affected children by January 1, 1997. For children whose re-determinations are made after July 1, 1997, benefits would end during the month following re-determination.

As noted above, SSI is principally funded by the federal government with a small state supplement. In New York City, the degree to which these changes will affect welfare rolls and result in City and State costs will depend on the percentage of children disqualified from SSI, as well as the percentage of these disqualified children whom the State and City subsequently absorb into the TANF program.

Using data from SSA (1996) and the New York State Department of Social Services (DSS, 1996), IBO estimates that 15,000 of the City's approximately 50,000 child SSI recipients will be removed from the program. Under the assumption that all of these children will be transferred to the TANF program, IBO projects the City's share of the total annual cost of providing TANF assistance to these children to be $8 million in 1998 and subsequent years.10

Legal Aliens

The new law restricts the eligibility of legal aliens for public benefits. While restrictions with respect to current aliens are discussed below in a later section, the Act's provisions concerning new immigrants will have a significant impact on TANF caseloads.

The new law determines eligibility by creating two categories of aliens: "qualified" and "not qualified." Qualified aliens may be any of the following: permanent residents; aliens granted asylum; refugees; aliens granted withholding of deportation; aliens paroled into the U.S. for at least one year; and aliens granted conditional entry into the country. All other aliens are considered not qualified.

Except for refugees, certain veterans and their dependents and a few other exemptions, new immigrants will be denied TANF assistance for at least their first five years in this country. In recent years, over 100,000 new immigrants have entered New York on an annual basis, and approximately 5 percent had qualified for AFDC cash assistance.11 A large majority of these aliens will be denied TANF benefits under provisions of the new law. Using Congressional Budget Office (CBO, 1996) estimates of the portion of new immigrants who would have received AFDC under the prior law but will now be ineligible for TANF, IBO projects a 6,800 person reduction in the TANF caseload for 1998 and increasing reductions for new immigrants in the following years.

Aggregate Cost of TANF Caseload Changes

Figure 4 reports the combined impact of the three provisions affecting TANF caseloads on the City budget due to changes in the projected caseloads over time. Combining these effects yields the net change in caseload projections, as indicated by the difference between lines 1 and 6 of Figure 4. On balance, IBO estimates that the TANF caseload will be greater than the prior-law AFDC caseload before fiscal year 2002, generating an additional cost to the City of $1 to -$4 million per year, as is shown on line 15 of Figure 4. For 2002, the TANF caseload is expected to be lower than the corresponding AFDC caseload under the moderate growth scenario. In 2002, this lower caseload will reduce the cost to the City by $45 million.

Funds From the TANF Block Grant

New York State has been notified by federal authorities that it will receive $2.359 billion in federal TANF block grant funds each year, through the year 2002.12 Given this amount, estimates of the total net fiscal cost to New York City of the Act depend in large part on the portion of the block grant that the State allocates to the City. Although a definite percentage has not yet been determined, preliminary indications suggest that the figure may be about 60 percent.

For the purposes of using a conservative funding assumption to derive estimates of the direct budgetary impact on New York City, IBO has based its analysis on this 60 percent allocation figure. Under this assumption, the City would have access to $1.415 billion in federal block grant funds each year, as shown in Figure 4. After accounting for the estimated funding for administrative costs, EAF and JOBS ($461 million in 1997 and $452 million in subsequent years), between $954 and $963 million per year through 2002 will be available for spending on income-maintenance programs.13 It is this amount that is comparable to the level of federal spending for AFDC prior to the establishment of TANF.

On the basis of the City's recent share of New York State AFDC caseload and spending, a strong case can be made that the New York City portion of the federal block grant to New York State should be significantly higher than 60 percent. According to the State DSS (1996a), for calendar year 1995, New York City accounted for 68 percent of State AFDC recipients in an average month. For the total year, 66 percent of all AFDC dollars spent in the State were expended in New York City. Each percentage point increase in the share of the State's block grant that its allocated to New York City represents an additional $24 million in federal funds available to the City for welfare programs. Thus, a percent distribution comparable to the City's share of AFDC caseload or spending would add as much as $190 million per year, significantly mitigating the fiscal burden of federal welfare reform.

It is also important to note that this question of the share of the TANF block grant available to fund local public assistance costs is an issue unique to New York State. No other state requires localities to bear a significant share of the non-federal costs of providing welfare assistance. New York City spending on programs that redistribute income has been a major factor underlying the City's structural deficit.

For the period through 2002, one important factor helping to significantly ease the fiscal impact on the City is that the amount of the new block grant is based on historic spending levels covering a period before local welfare reform efforts and economic recovery began to reduce local AFDC caseload levels. As a result, in the near term, the City's portion of the TANF block grant will exceed what would have been the federal share of cash assistance to recipients under prior law. Figure 4 compares the cost to New York City of the AFDC and TANF programs under the moderate caseload scenario and shows the State block grant funds allocation to the City for income- maintenance programs. For example, in the 1997 fiscal year, $954 million of the federal block grant expected to be allocated to New York City would be available for income- maintenance programs, as compared with the $829 million in federal funds needed for cash assistance given the projected TANF caseload. The $125 million difference is the amount of block grant money received by the City above and beyond expected needs for income- maintenance programs.

This amount is often referred to as a "surplus." or "windfall." In 1997 and 1998, the remaining funds after TANF assistance would be used to cover the costs of administering a larger work program for TANF recipients, the associated costs of providing day care for their children, or a part of the State's portion of cash assistance. By 1999, however, these additional costs are expected to exhaust and exceed the remaining funds. Due to the projection of small decreases in TANF caseloads from 1999 through 2001, the amount of block grant funds that remain after cash assistance under the moderate scenario is expected to increase slightly each year, reaching $162 million in fiscal year 2001 (line 14 of Figure 4). In 2002, the first year a significant number of recipients may be forced off the TANF rolls due to a five year cumulative limit, the amount of remaining funds is projected to rise to $254 million. Because the remaining funds reduce the estimated total net cost to the City, they are accounted for in Figure 9-a at the end of this section.

Work Requirements

The new law establishes work requirements for TANF recipients. Individuals may satisfy the work requirements through any of the following: unsubsidized or subsidized employment; on-the-job training and work experience programs; community service; 12 months of vocational training; or through the provision of child care services to individuals participating in community service. The limits placed on time spent in vocational training reflect a decreased emphasis on training and education.

The requirements contain two main elements affecting the number of recipients who must participate in work programs. First, states must meet minimum work percentage rates. The law sets the minimum rate for 1997 at 25 percent of all families receiving public assistance. This rate increases in 5 percent increments each year to 50 percent for 2002. States that fail to meet these rates will face fiscal penalties; five percent of the state's allocation is withheld if it fails to meet the minimum rate for the first year and penalties increase in subsequent years. States that successfully reduce their caseloads from the 1995 base are given credit towards their participation rate.14

Second, the new law mandates work activities for all recipients after two years of receiving benefits. Enforcement of this provision would effectively override the work rates discussed above and speed up the rate at which states would have to create work programs. Significantly, the law does not penalize states for non-compliance with this provision, and because it may prove difficult to implement, some states may choose not to enforce it.15

The decision whether or not to enforce the two year work rule will critically influence the administrative costs these work requirements would impose on the City and State. Enforcement of the two year work rule would vastly increase the number of recipients required to work beginning in 1999 as the first cohort reaches the time limit, thereby increasing City costs. Since it is possible that the State may not enforce the two year work requirement of the new federal law, each cost projection is presented with and without the two year rule.16 Estimates of the administrative costs associated with the new TANF work requirements are presented in Figure 5.

Using State and City caseload data (DSS, 1996b) and information from the Urban Institute (Pavetii, 1995), IBO estimates that the two year rule would require 60 percent of recipients to work in 1999 and 70 percent in 2000 and beyond (line 8 of Figure 5). If the two year rule is not enforced, only 22 percent of TANF recipients in 1999 would be required to work, rising by roughly 4.5 each year to 36 percent in 2002 (line 2 of Figure 5). Thus, by 2002 enforcement of the two year work rule would require 71,400 more work slots compared to the number needed without enforcement of the rule.

The City's current budget includes funds for work training programs for approximately 21,000 family heads. To avoid double counting, the administrative cost of these work slots is not considered an additional cost to the City of the shift to TANF. Thus, as indicated on lines 4 and 10 of Figure 5, the number of current work program cases is subtracted to obtain a net number of work program slots that must be funded under the new welfare law.

The IBO's estimate of the administrative costs of the new work program is derived from a survey of the costs of a variety of work experience programs throughout the country conducted by the Manpower Demonstration Research Corporation (Brock, et al., 1993). Using costs from a small-scale program in Chicago, adjusted for local transportation costs and inflation, IBO projects an annualized per worker cost of $2,250 in 1997. This estimate is quite conservative given that the costs of the Chicago program were at the low end of the range for the programs studied by MDRC.17 For subsequent years, the cost was adjusted for increases in the City's budgeted personal service costs. An additional adjustment was made to reflect the effect of the increase in the number of work slots required to meet the provisions of the Act. Although it is appropriate to assume that the marginal cost per work program participant would decline with modest increases in the number of workers, it is likely that marginal costs will actually increase due to diseconomies of scale given the very large number of new workers needed to meet the required participation rates (Brock, et al., 1993, p. 55). These adjustments yield projected per worker costs of $3,070 in 2002 if the two year rule is not enforced and $3,410 if it is enforced, reflecting the greater complexity for the City in carrying out a much larger work program.18

As shown in Figure 5 (lines 6 and 12), the total added cost of additional work program participants falls from 1997 to 1998 due to projected declines in the TANF caseload under the moderate caseload scenario even though the statutory required work participation rate has increased. This is because New York State will enjoy a credit against its required participation rate for having reduced its caseload since 1995. After 1998, when some TANF recipients reach the two year mark, enforcement of the two year rule significantly affects cost projections. If the two year work rule is enforced, the added cost to the City of administering the work requirement is projected to reach $478 million by 2001, $327 million more than without the two year rule. Because a certain number of recipients will reach the five year mark in 2002 and no longer receive benefits, the total number of TANF recipients in work programs is expected to fall, leading to a decrease in associated administrative costs under TANF.19 Still, with enforcement of the two year rule, the added cost is projected to be $424 million, $262 million more than without enforcement.

For the purposes of deriving an estimate of the new law's budgetary impact, IBO has put aside the question of whether it is indeed possible for the City's public and private sectors to provide the large number of work program slots required under the new federal law. So far, New York City's public sector has provided virtually all work program slots. It is not clear that government can absorb the projected increases in work program participation, ranging from 30,000 to 121,000 individuals in 1999, depending upon whether the two year work rule is enforced. Nor is it clear that the private sector will be able to generate enough new jobs to alleviate the problem. Led by the recent strength of financial service firms, the private sector has added over 140,000 jobs to the City's economy since the end of the national recession in November 1992, but many if not most of these positions require substantial skills that TANF recipients are not likely to possess.

Child Care Needs

As discussed in the previous section, the new welfare law requires states to meet rising work quotas for their TANF recipients, with particularly ambitious quotas beginning in 1999 if the two year rule is enforced. Implementing work programs large enough to comply with these quotas will force state and local governments to confront the problem of providing day care for the children of work program participants.

The IBO projects that by the year 2002, adjusted for already allocated new child care slots, between 33,000 and 84,000 full-time equivalent child care slots (40 hours per week) will need to be created, depending upon whether the two year work requirement is enforced.20 As shown in Figure 6, at an annual cost of $5,000 per slot in 1997 (adjusted for inflation in subsequent years), the additional cost per year of providing these new slots will range from $212 million to $599 million in 2002 (lines 7 and 13).

The current average cost per full time day care slot at the City's Agency for Child Development (ACD) is $6,500. This rate blends the cost of the various formal and informal day care options offered by ACD. City officials have publicly stated their intentions to increase the use of lower -cost informal day care by work program participants. Assuming that the City will have some success in lowering the overall average cost, IBO projects that child care costs will be $5,000 per slot per year in 1997. For subsequent years, the cost was adjusted by OMB's forecast of the City Consumer Price Index. As with work program administrative costs, an additional adjustment was made to reflect the effect of the large increase in the number of child care slots required to meet the work requirements of the Act. These adjustments yield projected child carecosts of $6,479 per slot by 2002 without enforcement of the two-year rule and $7,108 with enforcement.

Although the new law increases funding for child care, IBO projects that these new funds will be insufficient to meet the rising needs associated with the law's work requirements.21 Based on a formula supplied by the State Division of Budget, the City's share of the State's new child care block grant is expected to increase funding to support the TANF work rules, as shown in line 2 of Figure 6. New funds under the child care block grant are projected to equal $32 million in 1997 and rise to $70 million by 2002. Even when this pool of funding is accounted for, however, the projected additional costs to New York City net of block grant funds of the new child care slots to meet the work requirements in 2002 range from $142 million to $529 million, as shown on lines 8 and 14 of Figure 6. The $387 million difference between the two projections reflects the large increase in child care costs that would result from enforcement of the two year work rule.

Other Provisions Concerning Legal Aliens

In addition to the TANF restrictions on new immigrants discussed above, the new law also restricts eligibility of current aliens for SSI and food stamps and the eligibility of new aliens for SSI, food stamps and Medicaid. Because few not qualified aliens (see discussion above) are presently eligible for government benefits, provisions denying them aid will have little direct impact on the City budget.22 The more important distinction is between those qualified aliens exempted from the bans and those expected to feel its full impact. Unless they fall into an exempt category, current qualified aliens lose both their SSI and food stamp benefits. Examples of qualified aliens within the exempt group are refugees, asylees who have been in the U.S. for less than five years, certain veterans and their dependents, and permanent residents with 40 quarters of Social Security coverage.

Information obtained from CBO (1996) indicates that benefits will be halted for about three-quarters of those aliens currently receiving SSI, and that approximately 60 percent of aliens now on food stamps will see their payments cease. In addition, new immigrants who do not fall into an exempt category lose SSI, food stamps, and Medicaid benefits.

In order to gauge the fiscal implications of the provisions affecting immigrants, IBO assumed that the courts will interpret the State constitution to mandate that the State assist needy aliens who previously qualified for SSI or AFDC benefits and that the State will provide aid under the State- and City-financed Home Relief program.23 As shown in Figure 7, IBO estimates that the City's cost of caring for current and new immigrants will rise from $24 million in 1997 to $86 million by 2002 if the cost of Home Relief continues to be shared equally between the City and State.24 For current aliens, the estimates reflect the cost of moving individuals displaced from SSI onto Home Relief, as well as the assumption that the State will exercise its option to continue to claim federal Medicaid funding for aliens already here. For new aliens, the estimates account for the cost of moving individuals formerly eligible for AFDC or SSI onto Home Relief for both cash assistance and Medicaid.

Drug-Related Felons

A final direct budgetary impact relates to individuals convicted of drug-related felonies after the date of the law's enactment. Under the new law, such persons are barred for life from receiving TANF and food stamps, programs for which they formerly were eligible upon release from prison. States are permitted to override this new restriction. In either case, the families and children of drug-related felons would remain eligible for assistance.

This provision will have no practical effect for the next few years, since it will take some time for future convicted felons to commit crimes, be convicted, and serve out their sentences. Eventually, however, New York City will contain a growing pool of released prisoners ineligible for TANF or food stamp benefits.

According to the New York State Department of Correctional Services, the State held nearly 69,000 inmates in custody in 1995, 34 percent of whom were drug felons. During 1995, 10,000 convicted drug felons 70 percent from New York City entered the New York State prison system. The average length of stay was 27.5 months, but many released felons eventually return to prison. For example, of the more than 9,000 drug felons released in 1991, 38 percent were eventually recommitted. Based on this data, IBO estimates that the pool of released prisoners ineligible for TANF or food stamps could be as high as 20,000 individuals by the year 2002.

Because New York State currently provides coverage for drug-related felons, a decision by the State to continue federal coverage would leave the City's costs unchanged. Should the State elect instead to uphold the sanctions on drug felons, however, it is likely that the Home Relief caseload would increase and thus place a direct burden on the City budget. The population of drug felons released from prison may have particular characteristics that make them more likely to require public assistance, such as a high rate of HIV infection and difficulties in obtaining jobs because of their records. Because of the many uncertainties, the cost of the federal welfare changes with respect to drug felons is especially difficult to quantify, and, therefore, no estimate of their budgetary impact is provided here.

Total Net Fiscal Cost

Much of the projected total net cost to the New York City budget of recent federal welfare reform comes from the shifting of those becoming ineligible for TANF on to the City- and State-funded Home Relief program. Figure 8 summarizes the increase in caseload and corresponding cost of this shift. The City's share of Home Relief costs are projected to increase by an average of $57 to $64 million per year between 1997 and 2001, depending on whether the two year rule is enforced. Costs are expected to rise sharply in 2002, however, as the first cohort of TANF recipients reaching the five year eligibility limit is assumed to be transferred to Home Relief. If the two year rule is not enforced, IBO projects that New York City's share of Home Relief will increase by $246 million in 2002; with enforcement of the two year rule, the cost of Home Relief is expected to rise by $350 million.

The IBO estimates that 83,000 individuals will reach the five year limit and be transferred from TANF to Home Relief in 2002, causing City-funded public assistance spending to rise by $45 million in that year. However, the total cost of the shift to Home Relief will depend on a number of other factors, including the critical question of workfare requirements for Home Relief recipients. Confronted with rapidly rising caseloads, the State and City are likely to strengthen welfare-to-work programs for Home Relief. If welfare recipients who are shifted to Home Relief face the same work requirements as they previously faced in TANF, then Home Relief workfare administration and childcare costs would remain constant and the total increase in City-funded welfare spending would equal $45 million in 2002. This is the assumption behind IBO's cost projections. Less stringent workfare requirements in Home Relief could, of course, reduce the City's cost of transferring recipients from TANF but only insofar as the relaxed Home Relief work rules did not actually further swell Home Relief caseloads.

Figure 8 reports the additional Home Relief costs for cases removed from TANF, and SSI in all years including those reaching the five year limit in 2002.

The total net cost to the New York City budget, as computed under the assumption of moderate economic growth is presented in Figure 9-a. Because the two year rule concerning work program participation does not come into play until 1999, the net fiscal impact in the two prior years is the same regardless of the State's action. In these years, the total net fiscal impact on the City is modest. In 1997, the total net cost is $3 million, and in 1998 there is an $18 million surplus.

The 1998 surplus results from a projected dip in the TANF rolls during this period. With a reduction in TANF cases, the amount of TANF block grant funds available to offset gross costs of work program administration and child care increases. Moreover, there is actually a decrease in the number of TANF recipients required to participate in work programs, leading to associated savings in work program administrative and child care costs.

After 1998, the total net fiscal cost to the City will vary depending upon whether the two year rule is enforced by the State, though in both cases net costs rise over time. Without enforcement of the two -year work rule, annual net costs are projected to grow to $215 million and $269 million in 2001 and 2002, respectively, the last two years of our analysis. The decision to enforce the two year work rule would be far more costly to the City. By 2001, enforcement of the work rule would cause costs to increase by $1.039 billion more than four times the total net fiscal cost without enforcement. In 2002, the projected net cost is slightly lower, at $1.022 billion.

Direct Fiscal Impact Under Alternative Caseload Scenarios

Figures 9-b and 9-c summarize the Act's total net fiscal cost to New York City for the higher and lower caseload scenarios. The divergence of the two alternate scenarios from the moderate scenario pertains to 1999 and beyond, and thus the cost projections differ among the three scenarios for those years only.

Slow Growth, Higher Caseload Scenario

If the City's economy follows the slower growth scenario in the years after 1998, public assistance caseloads under both prior law programs (AFDC) and the new law (TANF) will be significantly greater than those in the moderate caseload scenario. These higher caseloads cause assistance payments to rise and also increase the costs of work program administration and child care (lines 3 and 12 of Figure 9-b). As a result, the City's costs are much greater under the higher caseload scenario.

Without enforcement of the two year work requirement, total costs to the City in 2002 are projected to be almost 15 times greater than in 1999. This compares with an eight-fold increase under the moderate growth scenario. The IBO estimates that the annual total net cost of the new law in a slowly growing economy would reach $392 million by 2001 rising to $563 million in 2002.

Enforcement is even more costly, especially in the first few years after the two year work requirement would come into play. In this case, the projected total net cost in 1999 is $733 million as compared with $38 million without enforcement. Estimated costs rise to $1.249 billion in 2002, $686 million greater than the comparable figure without enforcement.

If the slow economic growth underlying the higher caseload scenario actually does occur, the State may be able to tap into a $2 billion nationwide contingency fund provided in the Act. The State would have access to the fund if, either: 1) the unemployment rate during a three -month period was at least 6.5 percent and 110 percent of the rate for the corresponding period in either of the two preceding calendar years; or 2) the increase in its food stamp caseload was at least 10 percent over the 1995 level (adjusted for the impact of the law's immigrant and food stamp provisions on the food stamp caseload). It is entirely plausible that New York State could meet either of these criteria in the event of poor economic growth. State and City spending on cash assistance and work programs above the 1994 levels (not including child care) would be matched by federal contingency fund dollars.25

New York State has received 14.6 percent of the nationwide total amount of the first TANF block grant funds. If this percent is applied to the $2 billion contingency fund and it is assumed, as with the block grant, that 60 percent of federal money received by the State would get allocated to the City, as much as $175 million over five years from the contingency fund may become available to the City. But because funds available to New York would depend upon rival claims made by other states, IBO has not factored into the calculations presented in Figure 9-b any estimate of the amount of money that might become available.

Faster Growth, Lower Caseload Scenario

If the City's economy were to grow at a relatively fast pace in the next few years, yielding lower caseloads, the fiscal impact of reform would be lessened although it would remain significant. Strong economic growth would decrease TANF caseloads, and decreasing caseloads would in turn increase the size of the TANF surplus and reduce work program and child care costs. A decreasing caseload reduces the number of TANF recipients required to work in two ways: 1) it decreases the average caseload relative to that in 1995, which under the new law decreases the City's work quota in any given year; and 2) it lowers the base against which the annual work percentage rates are applied.

Figure 9-c presents IBO's estimates of the total net fiscal cost to the City of welfare reform under the lower caseload scenario (lines 9 and 18). These costs are significantly lower than under the moderate growth scenario, particularly for the case of no enforcement of the two year work requirement. Without enforcement, the projected net budgetary cost increases from $20 million in 1999 to $86 million in 2001. This latter figure compares to $215 million under the moderate caseload scenario and $392 million under the higher caseload scenario. As with the other scenarios, enforcement of the two year work requirement greatly increases work program and child care costs and thus the total net cost to the City of welfare reform. Under this scenario, annual net costs peak earlier than in the others, at $942 million in 2000. They then fall by over 16 percent, to $794 million in 2002.

The projections under the alternative scenarios provide a measure of the range of possible fiscal costs of the Act, with the range generally widening in the later years of the analysis. These projections are summarized in Figure 10. In 1999, for example, estimated net costs range from $20 million under the lower caseload scenario with no two year work requirement enforcement to $733 million under the higher caseload scenario with enforcement. By 2002, the range has expanded, from a $33 million total net cost without enforcement of the two year work rule under the lower caseload scenario to $1.249 billion under the higher caseload scenario with enforcement of the two year work rule.

Behavioral Changes and Caseload Scenarios

The incentives and sanctions contained in the new law including time limits, eligibility limits, and work requirements are designed to motivate recipients to leave the welfare system and secure employment in the private sector. If the new law were to succeed in changing the behavior of welfare recipients, the fiscal costs of implementing the law would be reduced. For example, under the moderate economic growth scenario, caseloads might continue to decline instead of remaining flat after 1998. Such a change would reduce City costs for income maintenance (in both the TANF and Home Relief programs), lower the percentage of recipients required to participate in work programs, and reduce overall workfare and child care costs.

If, under the moderate growth scenario, 5,000 fewer cases were opened and/or remained on the TANF rolls each year starting in 1998, 25,000 fewer adults would be on the City's welfare rolls by the year 2002. As a result, the total cost of welfare reform in New York City in 2002 would fall from $269 million to just $25 million (assuming that the two year work rule is not enforced). Such a change in behavior would result in cost reductions under the moderate economic growth scenario comparable to those under the assumption of faster growth.

While acknowledging the possibility that such a change in the behavior of welfare recipients could occur, it is also important to question the extent to which an additional 25,000 welfare recipients (above and beyond those who would otherwise leave the rolls) would succeed in securing private sector employment between 1998 and 2002. Under the moderate growth scenario, the New York City economy is forecast to add roughly 150,000 private sector jobs over the period. Such job growth, however, would only partly decrease the ranks of New York City's officially unemployed job seekers, who currently number close to 300,000. For this reason, the City's economy would be unlikely to fully absorb the added labor force participation resulting from declining TANF and Home Relief enrollment without the job creation associated with more rapid economic growth.

The Impact of State-City Fiscal Relations on the Cost of Welfare Reform

The decision to allocate 60 percent rather than 68 percent of New York State's TANF block grant to New York City will not only determine how much federal funding is available to offset the reform-related costs of work programs and child care in the City. The decision also will affect the share of the non-federal reform-related costs the City itself will have to bear.

Figure 10 reported the wide range of IBO's estimates of reform-related net fiscal costs to New York City under different caseload scenarios and different assumptions regarding the enforcement of the two year work rule. Similarly, the range of overall fiscal costs for TANF, Home Relief, and related work programs and child care is great. In 2002, for example, estimated overall costs to the City budget range from a low of $593 million under the low caseload scenario without enforcement of the two year work requirement to a high of $1.93 billion under the high caseload scenario with the two year work rule. At the same time, New York State's estimated overall costs for TANF, Home Relief, and ancillary programs significantly affected by reform in New York City varies much more narrowly, ranging from approximately $688 million under the low caseload scenario to $815 million under the high caseload scenario in 2002.

Under any of the scenarios detailed in this report, the total cost to the federal government of welfare spending in the City will remain fixed at $963 million from 1998 to 2002. The contrast of fixed federal costs and varying non-federal costs reveals that the Act has, as was intended, shifted the fiscal risks and possible rewards of welfare reform from the federal level to the states. But the contrast of narrowly varying State costs and widely varying City costs also shows that in New York most of the fiscal risk has been shifted even further, from the state level to the local government level.

Under the six alternative scenarios examined above, the State's overall estimated costs for TANF and Home Relief in New York City vary by no more than $127 million or 8.5 percent above or below the costs associated with the moderate caseload. The range of estimated overall costs for New York City, however, is much greater a remarkable $1.25 billion in 2002, nearly ten times the projected range of the State's overall costs.

Furthermore, IBO's projections of overall costs of providing welfare indicate that in all but one of the six scenarios detailed in this report, the estimated overall City-funded costs of TANF, Home Relief, and reform-related work programs and child care exceed the comparable costs for the State. Only under the most favorable scenario of a low level of caseloads with no enforcement of the two year work rule does the projected cost to the City fall below the State's overall costs for New York City ($593 million and $688 million in 2002, respectively). Under every other scenario, City Hall bears a larger share of the City's overall welfare costs than Albany. Indeed, in four of the six scenarios, the City also bears a larger share of welfare costs for City programs than the federal government, including one scenario where the City's costs are estimated to be larger than the federal and State costs combined.

In general, the non-federal costs of welfare are no longer split evenly between the City and State. For example, under the moderate growth caseload scenario with no enforcement of the two year work rule, IBO estimates that the State will fund only 46 percent of the non-federal costs of welfare in 2002. Under the high caseload scenario, with no enforcement, the State would pay only 39 percent of a significantly greater non-federal burden.

These disparities in the City and State burdens of welfare spending would be substantially reduced, and the City's fiscal risk under the Personal Responsibility and Work Opportunity Reconciliation Act would be significantly eased if New York City were to receive a TANF allocation in line with its 68 percent share of total AFDC caseload.

Welfare Reform: A Longer View

The new welfare law's most significant fiscal impacts are likely to occur in the years 2002 and beyond. Over this period, the total fiscal cost to New York City of providing welfare could increase sharply, while the size of the TANF block grant the State receives from the federal government is likely to decline. Moreover, changing welfare from an entitlement program to a block grant shifts the burden of coping with economic downturns from the federal government to states and localities; the longer the time horizon, the more likely that a downturn will confront the State and the City with particularly difficult choices.

The full impact of the lifetime limit on welfare participation only begins to be felt in 2002, when the first cohort of recipients reaches the five year limit. As previously discussed, IBO projects that by 2002, roughly 31 percent of TANF recipients will reach the time limit, considerably above the 20 percent that the State is allowed (but not required) to exempt for reasons of hardship. However, New York State's constitutional requirement of support for the needy makes it likely that individuals removed from the TANF rolls would be placed on Home Relief. Since, New York City is responsible for 50 percent of Home Relief spending but only 25 percent of TANF, however, the City's annual cost for each welfare recipient reaching the limit would double from about $550 to $1,100.

The decline in TANF caseloads as some recipients reach the five year limit in the year 2002 could mark just the beginning of a larger long term decline. Unless the five year limit is allowed to expire after 2002, it will force the removal of progressively larger numbers of TANF recipients over subsequent years. It will also reduce TANF caseloads by blocking new applications for public assistance by persons who have previously accumulated five years of TANF benefits. Assuming the Act remains unchanged, these post-2002 reductions will continue until the caseload is reduced to less than half its current size. Indeed, this follows inevitably from the fact that well over half of the current caseload consists of recipients who have accumulated or will accumulate more than five years of public assistance.

The bulk of this massive caseload reduction would occur just over the 2002 horizon. The IBO projects that about 100,000 more recipients would be forced off TANF in 2003, and another 15,000 denied entry by the five year limitation. In 2004, an additional 120,000 would be removed and 25,000 more denied entry. In 2005, the numbers rise to 155,000 taken off the rolls and 30,000 denied entry. The total TANF caseload would be reduced to under 300,000 individuals by the year 2005 450,000 fewer than it would have been under the continuation of AFDC. In 2006 and subsequent years, IBO projects that the number of individuals removed from the TANF rolls or blocked from re-applying due to the five year limit would finally begin to decline.

As in 2002, these reductions in TANF caseloads could cause City welfare costs to rise. With the New York City economy able to absorb only a small fraction of the individuals removed from TANF and given the State's constitutional mandate to care for the needy, individuals removed from TANF are likely to continue to shift to Home Relief. In 2005, the total City cost of this shift could exceed $225 million assuming that the per case costs of public assistance remain constant in nominal terms. By then, however, even moderate inflation would have lowered the real dollar value of public assistance to less than 80 percent its 1997 level. In the past, public assistance benefits have seldom been fully adjusted for inflation and adjustments have generally occurred with considerable lags. The longer such adjustments are postponed, however, the less credible is the assumption that current per case costs will not rise. Restoring just half the erosion of benefits due to inflation would add some $25 million to the City's cost of shifting welfare recipients from TANF to Home Relief in the year 2005.

While some individuals forced off the TANF rolls might look for better prospects in other states or regions with more robust economies, the changes triggered by welfare reform might well prompt a migration of benefit-seekers into New York, thereby swelling City welfare expenditures in the long run. Economic research suggests that the switch from federal matching grants under AFDC to block grants under TANF will lead most states to lower welfare benefits and cause the differential in benefits across states to widen (Chernick, 1996; Gold and Lowenstein, 1996). New York is likely to remain among the states with the highest benefit levels, both because of its constitutional mandate to assist the needy and because the increase in the price of providing welfare benefits is smaller for New York than for most other states.26 Although econometric evidence on whether differentials in benefit levels have spurred welfare recipients to migrate is ambiguous, New York State may well become a more attractive destination for the needy as benefit differentials widen.27 Assuming only a small fraction of welfare recipients move from low- to high-benefit states in any given year, the impact on New York City and State caseloads would be substantial over a number of years.

Even if the cost of providing welfare does rise in 2002 and beyond, it is quite possible that New York State's TANF block grant will decrease. Under current law, average welfare caseloads for the period 1992 to 1994 are used to determine New York State's share of federal TANF funds. The State and City are now benefiting from this determination, because the level of the block grant does not reflect the recent sharp reduction in caseloads that have resulted from the NYC WAY program. Welfare reform legislation is scheduled for renewal in 2002, however. If the formula for determining each state's share of TANF block grants remains the same and if total U.S. funding for the program does not rise, then New York State's grant will decline.

Finally, expanding the time horizon beyond the year 2002 increases the probability of an economic downturn. New York City was hit particularly hard during the most recent recession, with AFDC caseloads rising at an average annual rate of 6.6 percent from 1990 to 1992. Under the new welfare law, however, the size of New York State's TANF block grant would not be affected by a recession-induced increase in caseloads.28 As a result, a downturn would force the City and State to make difficult choices between decreasing expenditures as needs increase and/or raising revenues as ability to pay declines.

Conclusion

The Act's impact on the New York City budget will ultimately depend on the decisions of State and City officials, the performance of the local economy, and the behavior of thousands of individuals in response to the new system. These uncertainties coupled with the newness of the program make estimating the fiscal impact of the new law particularly difficult.

Assuming that the City receives 60 percent of the State's TANF block grant, IBO projects that the Act will have little impact on the City's budget over the next two years. Beyond 1998, costs could rise significantly depending on caseload. The timing of these costs depends on whether or not the State chooses to enforce the two year work rule or to phase in federal work requirements more gradually. Mandating welfare recipients to work has the effect of significantly increasing the City's cost of administering work programs and providing child care.

Under the moderate caseload scenario, IBO projects that the City's cost of providing welfare would increase by $33 million in 1999 growing to $269 million by 2002. If the two year work rule is enforced, these additional costs would total $723 million in 1999 and roughly $1 billion annually through 2002.

Under the higher caseload scenario, additional costs would increase by $38 million in 1999 growing to $563 million by 2002. If the two year work rule is enforced, estimated additional costs would total $733 million in 1999 and more than $1.2 billion by 2002. In contrast, lower caseloads could result from some combination of more robust economic growth and the law's potential success in moving welfare recipients off the rolls and into private sector employment. Under a lower caseload scenario, the City could expect to incur additional costs of less than $100 million annually. If the two year work rule is enforced, additional costs would range from $700 million to $942 million over the 1999-2002 period.

Further, it is important to emphasize that the State's decision on how the TANF block grant will be shared is a critical factor in determining the cost to New York City. The City currently accounts for 68 percent of all AFDC recipients. If New York City were to receive 68 percent of the TANF block grant as opposed to the 60 percent assumed in the analysis then the City would receive an additional $190 million each year thereby dramatically reducing expected costs.

Given the fact that much of the cost discussed in this report results from worker placement and related child care costs, the price tag of the new law could be substantially lowered if the local economy is able to grow at a rate sufficient to create and sustain employment for the tens of thousands of welfare recipients in need of work. Although future economic performance is unknown, it is clear that the new welfare law's most significant fiscal impacts will be felt in the year 2002 and beyond, as a greater share of public assistance is borne by the City and State. Given existing and projected budgetary difficulties, such increased costs will make it ever more difficult for the City to meet the basic needs of its poorest citizens.

Appendix A

Food Stamps

The Act sharply reduces the level of food stamp assistance in the City by eliminating some categories of recipients from the program and lowering benefits for those who retain their eligibility. The IBO estimates that overall food stamp spending will be reduced by between $400 and $500 million per year, or approximately 35 percent of current spending levels. The food stamp program is funded entirely by the federal government. There is no State or City counterpart. Therefore, reductions in the program have no direct fiscal impact on the City's budget. However, there will likely be indirect fiscal impacts, as well as the certain direct impacts on individuals. Person losing assistance will be directly affected, and this in turn is likely to indirectly impact the New York City budget.

The Act changes the food stamp program in the following ways. First, it removes many current legal aliens from eligibility for any benefits, as well as many aliens arriving in the future. Based on data from the Census, IBO estimates that there are currently 206,000 legal aliens receiving food stamps in the city. Assuming that the same share of newly arriving aliens would have qualified for food stamps, IBO estimates that each year approximately 15,000 new aliens would have joined the ranks of food stamp beneficiaries. Under the new law, eligibility is based on the same qualification categories described above in the section on TANF caseloads. The CBO has estimated that the new policy will eliminate 60 percent of legal aliens previously eligible for food stamps nationwide. As shown in Figure 11, based on these estimates IBO projects that nearly 130,000 legal aliens will lose their food benefits by 1998 when the new provisions are fully phased in. This provision will account for $121 million of the total reductions in that year.

Second, the Act limits the tenure of some food stamp recipients. Childless individuals who are not working or enrolled in job-training programs are limited to three months of food stamp assistance within a thirty-six month period, with one additional three-month extension for individuals who obtain jobs and then subsequently lose them.

Third, the Act expands the income base used in determining eligibility for food stamps. For example, state and local energy assistance will now be included as income, the age of children whose income must be included in household income is lowered from 22 to 18, and the shelter deduction cap is frozen.

Fourth, the new law reduces the value of food stamp benefits actually paid to recipients. The benefit level is reduced from 103 percent to 100 percent of the Thrifty Food Plan, a market index designed to measure the cost of food for an average family, and the legislation repeals a scheduled increase in the food stamp minimum.

Finally, as discussed in this report, the new law eliminates drug felons from food stamp eligibility.

Using data from the U.S. Department of Agriculture, IBO estimates that the total reduction in food stamp spending attributable to these last four provisions will be between $300 and $400 million per year once they are fully phased in. (See line 5 of Figure 11.)

By their nature, indirect fiscal impacts are difficult to fully inventory, let alone quantify. Nevertheless, some observations are possible.

First of all, the reduction in overall spending will withdraw between $400 and $500 million of subsidized purchasing power from the local economy, most of which had been concentrated in low-income areas of the city.29 This withdrawal, accounting for more than a third of current food stamp spending in the city, could threaten the profitability of small merchants and businesses, particularly in poorer neighborhoods where food stamp subsidies and welfare payments account for much of the disposable income of residents. This in turn would lower business tax revenues and commercial property values in those areas.

Secondly, to the extent that individuals facing eliminated or lowered food stamp benefits reduce the quantity or quality of the food they eat, they are more likely to encounter health problems related to nutritional deficiencies and their children may face diet-related developmental problems. Given that most food stamp beneficiaries are also eligible for Medicaid and many rely on municipal hospitals, the costs of dealing with such health problems would be borne in part by the City. Moreover, children whose early development is impaired can be expected to increase the enrollment in the public schools' special education classes which have much higher per pupil costs than do classes for general education.

Finally, it is possible that individuals denied benefits, particularly those with prior drug felony convictions, would be more likely to commit crimes. Increased crime has many attendant costs for the City, ranging from increased policing to the need for more prisons, as well as a perception of lower personal safety and diminished quality of life.

Aside from the fiscal impacts, whether direct or indirect, the changes to the food stamp program will have very real direct impacts on the individuals involved. Many of those eliminated from the program will face substantial reductions in their disposable income. Those with reduced benefits will have smaller but still measurable reductions. The City might attempt to offset this impact on recipients and prior recipients by creating a new food subsidy program, but as indicated in Figure 11, the cost of ameliorating these cutbacks would be considerable.


Footnotes
1The social and economic impact of welfare reform on the lives of New Yorkers is generally beyond the data currently available and the scope of this report. One major exception is the reduction in food stamp benefits, which is expected to have little direct fiscal impact, but is likely to significantly affect individuals receiving assistance and the economies of the neighborhoods in which they live. (See Appendix A.)
2Although the improving local economy was also a factor, the vast majority of the caseload decline was attributable to NYC WAY.
3This analysis assumes that the State will not reduce Home Relief grant levels, reduce eligibility, or place time limits on Home Relief recipients. Such actions could increase stress on other parts of the social service system - including transitional and subsidized permanent housing, foster care and protective services for children - which in turn could result in the need for additional City funds. These potential fiscal costs are not accounted for in this report.
4Several of the issues that TANF implementation legislation could address, including those related to the cumulative time limit of benefits and work requirements, are mentioned below. Because all of the details of implementing legislation are impossible to predict, IBO has focused on those issues likely to have the greatest impact on the fiscal cost to the City.
5IBO's analysis of welfare caseloads and macroeconomic data indicates that between 1984 and 1993, growth in AFDC caseloads had a strong positive correlation (0.84 percent) with the rate of unemployment and negative correlations with the growth rates of employment and real gross city product (0.51 percent and 0.45 percent, respectively). Moreover, a 1 percentage point decline in employment growth leads to an increase of about 1.13 percentage points in the growth rate of caseloads over the same period.
6The Act requireseach state to maintain state and local spending at levels that equal at least 75 percent of 1994 spending for the programs covered under the TANF block grant. For states that fail to meet the work participation requirements, the maintenance of effort standard rises to 80 percent. For the City portion of total statewide spending, the maintenance of effort requirement would be achieved under any of the scenarios discussed in this report.
7Criteria for determining hardship would need to be defined in the State's implementation legislation.
8In effect, City income-maintenance spending on the population of TANF clients who reach the five-year limit in 2002 simply doubles because it is assumed the City will be responsible for 50 percent of Home Relief spending yet only 25 percent of TANF spending. The calculation assumes that the per person benefit under the TANF and Home Relief programs would be equal for those being shifted. Currently, per person benefits of Home Relief recipients are greater than those for AFDC recipients because most Home Relief recipients are childless adults. The IBO's analysis assumes that the Home Relief population will become more heterogeneous, and that benefits per person would become lower.
9It may be possible for states to avoid the effect of these time limits altogether by juggling federal and state funds. Since the block grant system does not require states to match federal TANF spending, a state may be able to spend all of its federal block grant funds on newer TANF recipients while spending state funds on those who are over the time limit.
10To the extent that children transferred from SSI to TANF are physically or mentally impaired, the cost of meeting their special needs is likely to exceed that of other children receiving TANF assistance. For the purposes of this analysis, however, it is assumed that the per case cost of assisting these former SSI recipients is equal to the average for all children in the TANF program.
11Another 3 percent of recently arrived aliens qualify for cash assistance under SSI.
12After two months of analysis by the Governor's bipartisan Welfare Reform Task Force, the Governor submitted an interim plan to the federal government on October 16, 1996. Filing an interim plan, subject to change to reflect State legislative action, enables the State to remain eligible for the entire amount of the TANF block grant while the Governor and the Legislature achieve consensus on a legislative package to implement the Act.
13Programs that have been funded by EAF and JOBS, will now be funded by a portion of the TANF federal block grant. The EAF has been used to fund a large portion of the family shelter system as well as parts of the City's foster care and child abuse prevention programs. The block granting of these funds means that no additional federal dollars will be available to cover any future increases in demand for these programs.
14Credit is not given for caseload reduction attributable to revisions in the new federal welfare law.
15In its TANF implementation legislation, the State may provide for the waiver from the two year work rule for persons living in areas with an unemployment rate greater than 10 percent or where the Governor certifies there are "insufficient jobs." The State also has the option of exempting caretakers for children under one year of age from TANF work requirements for a total of 12 months on assistance, though IBO does not assume this policy option in its analysis.
16The Governor's interim plan does, however, indicate an intent to enforce the two year rule.
17IBO's conservative cost estimate is partially offset by a corresponding assumption that the economic value of the work program's output is negligible. To the extent that work program participants enable the City to produce its current level of services at a lower cost, these savings would ultimately reduce the cost to the City of work program participation.
18This analysis of work requirement costs does not include costs that may be imposed by existing state prevailing wage laws for welfare, home relief and JOBS work programs. Existing statutes such as 336-c(2)(b) and 336-h(3) of Social Service Law currently require work program participants to work no more than the hours necessary to exhaust the benefit level at the rate of the higher of the minimum or prevailing wage. The combined effect of the statutes and the act's specific minimum-hour work requirements could require significant increases of benefit levels with non-federal funds to avoid sanctions for failure to satisfy the work participation requirements.
19However, it is assumed that those shifted to Home Relief will be subject to similar work requirements. The costs of administering the comparable Home Relief program are discussed below.
20This projection is derived from information on the number and age distribution of children of current AFDC recipients (DSS, 1996b), the number of hours that TANF recipients are required to work, and the estimates of TANF caseloads. There is an average of 1.85 children per AFDC case. The number of hours that TANF participants are required to work depends on the ages of any children they may have, and thus the IBO's estimate of child care needs takes into account the age distribution of children.
21The new federal law combines mandatory and discretionary funding streams with the existing Child Care and Development Block Grant and eliminates transitional child care and JOBS employment-related child care.
22An exception is the class of aliens known as Persons Residing Under Color of Law (PRUCOL). These individuals, who are known to the government, fall in a middle area between legal and illegal status, and in some cases, they have been eligible for SSI and Medicaid. Their eligibility will cease under the new law.
23Absorbing legal aliens, children and other groups made ineligible under the Act into Home Relief will change the demographic composition of that program significantly.
24Drawing on information from the Census Bureau, the Immigration and Naturalization Service (1996) and the Department of City Planning, IBO estimates that there are currently 1.5 million non-citizens in New York City, and that 8 percent of recent immigrants are receiving either AFDC or SSI. The CBO (1996) estimates that 25 percent of aliens nationwide would remain eligible for SSI and 40 percent would remain eligible for food stamps, and this analysis assumes that the same proportions would apply for aliens in New York City.
25Eligibility for contingency funds requires the State to maintain spending at 100 percent of 1995 levels.
26Although the switch from federal matching to block grants will increase the price of providing benefits in every state, this price effect will be smallest in high-income, low federal matching rate states like New York.
27The interim plan filed by the Governor takes advantage of the federal option to fund new State residents for six months at their lower state welfare benefit levels.
28As noted above, although there is a contingency fund for economic downturns under the new law, the amount of funds potentially available to New York would be limited and depend upon competing claims by other states.
29We are assuming that current food stamp eligibility criteria are effective in determining whether an applicant requires the benefits in order to maintain a basic nutritious diet and that therefore, facing lower or eliminated benefits, (former) recipients will be forced either to lower their food purchases, or to shift spending from other categories such as clothing or shelter in order to sustain a healthy diet.

References

Brock, Thomas, David Butler and David Long. 1993. "Unpaid Work Experience for Welfare Recipients: Findings and Lessons from MDRC Research." Manpower Demonstration and Research Corporation, New York, N.Y. (September).

Chernick, Howard. 1996. "Fiscal Effects of Block Grants for the Needy: An Interpretation of the Evidence." Hunter College, City University of New York, New York, N.Y. Draft (August).

Congressional Budget Office. 1996. "Federal Budgetary Implications of H.R. 3734, The Personal Responsibility and Work Opportunity Reconciliation Act of 1996." Washington, D.C. (August).

Gold, Steven D. and Ronnie Lowenstein. 1996. "Federal Aid Cuts, the Balanced Budget Amendment, and Block Grants: Impacts on the States." The National Tax Journal, Papers and Proceedings Volume of the National Tax Association. (Forthcoming.)

New York State Department of Social Services. 1996a. "Social Statistics." Albany, N.Y., various issues.

New York State Department of Social Services. 1996b. "Annual Report and Statistical Supplement, 1993." Albany, N.Y.

Pavetti, LaDonna. 1995. "Questions and Answers on Welfare Dynamics." The Urban Institute, Washington, D.C. (September).

U.S. Immigration and Naturalization Service. 1996. "Immigration to the United States in Fiscal Year 1995." Washington, D.C. (August).


Please direct any inquiries to:
Independent Budget Office
110 William Street, 14th floor
New York, New York 10038
(212) 442-0632 - voice
(212) 442-0350 - fax

Please see this Addendum. Added 3/1998.