Statement of Harry J. Holzer, Professor of Public Policy,
Georgetown University, and Visiting Fellow, Urban Institute

Testimony Before the Subcommittee on Human Resources
of the House Committee on Ways and Means

Hearing on "Rainy Day" and Other Special TANF Funds

April 26, 2001

The Personal Responsibility and Work Opportunity Reconciliation Act of 1996 created a fixed level of block grant funding to states to pay for the Temporary Assistance to Needy Families (TANF) program. Since the size of the block grant itself does not vary in response to changing economic circumstances (such as a downturn) that might create greater caseloads, the 1996 legislation also established a contingency fund of $2 billion to provide states with an additional source of revenue in case of unforeseen developments.

The contingency fund is currently set to expire in five months, at the end of FY2001. Should it be reauthorized? Despite the unspent TANF surpluses that have been accumulated by many states, does there continue to be a need for an additional federal funding source for TANF in the event of an economic downturn? If so, what conditions should trigger the availability of these funds to the states, and how large should the fund be?

Below I argue that the contingency fund should be reauthorized, with a somewhat different set of triggers than were established in 1996, and with an uncapped level of funds available to states in need. The TANF supplemental grants to 17 relatively poor or rapidly growing states, due to expire this year, should be extended for at least a year as well.

An Economic Downturn Could Increase TANF Caseloads Substantially

A recession is a nationwide economic contraction that would be felt in virtually all states, though to varying degrees. A recession could be relatively mild, in which case the national unemployment rate might rise by just 2-3 percentage points (as in 1990-92); or it could be more severe, in which case the increase in the national rate could be 4-5 points (as in 1981-82) or more. During the most recent recession, unemployment rates rose by a percentage point or less (in annual terms) in about a third of the states, but they also rose by 4 points or more in many others.

If a recession occurs, job availability for those seeking employment will decline, and unemployment will rise among all groups, especially less-educated workers and those with limited skills or labor market experience. The increased unemployment would be what economists often refer to as involuntary unemployment - i.e., unemployment reflecting economic conditions rather than the choices made by workers themselves about whether or not they want to work. Particularly, employment will decline significantly among disadvantaged women, and the welfare caseloads will rise (Holzer, 2000).

But the sizes of these prospective caseload increases are uncertain, since we cannot predict the severity of the recession, or how states will respond (in terms of the ease with which they allow women to return to the welfare rolls). One recent study by the President's Council of Economic Advisers predicts that the rolls will rise by 5-7 percentage points for every one-point increase in the national unemployment rate. This estimate is based on a statistical analysis of the extent to which a stronger economy is responsible for the caseload decline of over 50% since 1994. Other estimates suggest that the strong economy of the late 1990's was an even larger factor in the caseload decline, and therefore that the rolls could increase by as much as 8-10 percentage points for each percentage-point increase in unemployment.⁽¹⁾ Again, the increases will almost certainly be larger in some states than others. Furthermore, these increases may persist for several years, since changes in welfare rolls tend to lag behind changes in the unemployment rate (which themselves can persist for several years during and after a downturn).

The Need for a Federal Response to Limit the Financial Strains on States

During recessions, state revenues decline while their expenditures on social services and support (such as TANF) generally rise. But, since most states need to balance their budgets annually, these economic fluctuations can create serious financial strains. Indeed, many states will face an incentive to limit the extent to which they allow low-income women (and men) to return to the rolls, and may use various administrative means of diverting them from doing so. If this occurs to a large extent, the hardships experienced by low-income women could be very substantial.

Of course, many states have been accumulating unspent TANF surpluses since 1996, as their caseloads have declined substantially while their federal block grants have been fixed (at $16.5B annually). The most recent numbers available (through the end of FY2000) show cumulative unobligated surpluses of under $3B nationwide, and unliquidated obligations of about $5B (Lazere, 2001). As the latter have been at least nominally obligated to various expenditure categories, their availability in case of a downturn will be uncertain. Furthermore, the sizes of the surpluses vary dramatically across states. For instance, several of the states with the largest caseloads (such as California, Pennsylvania and Florida) have virtually no unobligated surpluses and fairly small unliquidated sums (measured as percentages of TANF grants in any year).

In addition, state spending on related matters, such as child care assistance, have risen substantially in the past few years. Thus, a large majority of states spent virtually their entire TANF block grant in FY2000, and a dozen states even dipped into their surpluses to finance these activities. In the event of a downturn, states will likely face unpleasant tradeoffs between continuing these supports for working families v. reducing them to fund rising caseloads.

In sum, many states would face serious financial strain in the event of a downturn, and choices between continuing spending on popular and legitimate work-support activities v. increased spending on a rising caseload. Though the magnitude of the hardship will depend on the severity of the recession, its distribution across states, and their accumulated surpluses, a strong case can certainly be made for continuing the option of federal assistance to many of these states in need.

Are the Earlier Triggers for Receiving Aid Still Valid?

According to the 1996 law, states would have to meet several criteria before they could qualify for contingency funding during a downturn. For instance, the state's unemployment rate would have to be at a level of 6.5% or more and at least 10% higher than it was in either of the previous 2 years. Alternatively, states could qualify based on increases in their food stamp caseloads, as long as the caseloads were at least 10% higher than in FY 1994 or 1995 (adjusting for changes in eligibility enacted as part of the 1996 law). In addition, states must be spending at least as much on TANF as they were in FY1994 of their own money (Falk, 2001).

Do these triggers make sense in the current environment? All of these conditions have become more difficult to meet than they might have been in 1996, due to either declining unemployment rates or declining TANF and Food Stamp caseloads nationwide. For instance, the required 6.5% state unemployment rate was chosen in a year (1996) when the nation's unemployment rate was 5.4% - which was virtually identical to the national unemployment rate in 1989, right before the previous recession occurred. During the 1990-92 recession, over 30 states would have met the unemployment rate requirements for access to the contingency fund (Falk, 2001), as they represented average increases of only 1.2 percentage points above the existing national rate.

However, the nation's unemployment rate has averaged just about 4% during the past 2-3 years. If a recession of the same magnitude as that in 1990-92 were to now occur, it is likely that most states would fail to meet the 6.5% threshold. For instance, during the year 1999, 25 states had annual average unemployment rates below 4%, and 10 states had rates below 3%. The former group would need to have unemployment rates increase by well over 50%, and the latter would need increases of well over 100%, in order to qualify for contingency funding. Unemployment increases smaller than these would still generate hardship for these states and for the low-income women seeking employment there, but the states would not qualify for contingency fund assistance under those circumstances.

Meeting the food stamp and expenditure-level criteria would be more difficult than earlier as well. Food stamp caseloads have declined by roughly 40% since the mid-1990's, vastly more than was expected on the basis of eligibility changes. While these will no doubt rise again during the next downturn, it seems highly unlikely that they will increase by 80% above their current levels (which would be required to reach 110% of the FY1994 or 1995 caseloads). And, with most states currently spending 75-80% of what they were spending out of their own funds in the pre-1996 period, increasing their expenditures back up to 100% (given the reallocations of money to other areas that have occurred since that time) would be difficult as well for many states.

Under these circumstances, what kinds of changes in the triggers would be most appropriate? One possibility is to have a lower threshold level for state unemployment rates - such as 5% - perhaps combined with a more stringent threshold for increases in the rate (e.g., 20% rather than 10%). A 5% threshold level would constitute an increase in average unemployment relative to the recent 4% benchmark of roughly the same magnitude as a 6.5% threshold constituted relative to the 5.4% unemployment rate of 1996. However, since a 5% threshold may still be very difficult for states to meet that currently have unemployment below 3%, an alternative threshold based only on increases in the state rate (such as 50%) might be appropriate as well.

Similarly, a threshold for increases in the Food Stamp caseload that is more stringent than the current required increase, but calculated relative to the current base rather than the FY1994-95 base, might make sense as well. For instance, a required increase of 20-30% over current levels should be large enough to exceed any increases that might occur for other reasons (such as improved outreach) and would therefore capture labor market difficulties for unskilled workers that may not be fully reflected in the overall state unemployment rate. Finally, a state spending requirement of 75-80% of FY1994 levels (the same as current "maintenance of effort" requirements under TANF) would be a more realistic reflection of spending levels and obligations for most states than a return to 100% of the FY1994 levels.

Should the Fund be Capped at $2B?

In the event of a recession, it is possible that a $2B fund, in addition to accumulated TANF surpluses and access to the limited TANF loan fund, might be sufficient to meet state needs. On the other hand, if the downturn is relatively more severe and lasts for a number of years, it is also possible that this level of funding will prove grossly insufficient to meet the extra financial needs of states.

For instance, the $2B funding level constitutes less than one-eighth of one year's TANF block grant, and well under that percentage of total spending for cash and related assistance at the state level. Yet, in a recession, caseloads could rise by 10-30% (using the Council of Economic Advisers' estimates) or by significantly more (according to other estimates), and could persist for several years. Such increases could easily swamp the $2B contingency fund and available TANF surpluses, especially in the large states that have accumulated little to date in the way of surpluses.

A more sensible approach would probably be to leave the total amount of the contingency fund uncapped. Given that the funding mechanism in the current law already requires (and should continue to require) a match from the states, there are incentives in place for the states to economize on their use of these funds. An uncapped contingency fund would ensure sufficient availability to meet the legitimate needs of states without encouraging wasteful spending on their part.

Other Concerns Regarding TANF during a Downturn

While an uncapped contingency fund with appropriate trigger mechanisms would help alleviate the financial difficulties faced by states during an economic downturn, other concerns would remain, especially about the well-being of low-income women who would have difficulty gaining employment at that time.

At least a few of these concerns are as follows (Holzer, 2000):

1) Many women who have been on TANF may be ineligible for further funds, if they have reached their five-year lifetime limits; and most of them will also not qualify for Unemployment Insurance, either because they have too little work experience or because their reasons for leaving their previous jobs make them ineligible;

2) States will have more difficulty meeting their caseload work requirements under TANF during a recession. Many states have met the current requirements through caseload reductions rather than work among current recipients. Both will be harder to achieve when the economy turns downward.

While resolving these issues lies outside the scope of the current hearings, they need to be addressed to ensure the existence of a realistic safety net during a downturn for low-income women. Some options that we might want to consider to alleviate these potential problems include:

· Reforms in the Unemployment Insurance system to ensure higher rates of coverage for low-income women with at least some labor market experience;

· Reforms in TANF (perhaps implemented during reauthorization in 2002) that would temporarily suspend time limits in states with high unemployment and/or enable states to count a wider range of education and training activities towards their work requirements; and

· The provision of technical assistance and/or funding to states that want to implement Community Service Jobs programs to welfare recipients during a downturn, as a way of meeting work requirements and providing meaningful work experience to women who will have more difficulty finding it in the private sector.

Other Funding Issues: Supplemental Grants

Though my testimony today has focused primarily on funds that would be available to states during a downturn, a separate issue involves the status of TANF Supplemental Grants to 17 states that are either relatively poor or rapidly growing. Like the contingency fund, the supplemental grants are also due to expire this year; and their expiration would entail cuts in TANF block grants of roughly 10% to most of the states receiving these funds.

While these issues are discussed in greater detail elsewhere (e.g., Primus and Lazere, 2001), the supplemental grants should be extended by at least one more year, at a cost of roughly $400M, to ensure that current funding streams remain in effect until TANF reauthorization that will occur in FY2002.

Conclusion

At this point, it is impossible to know whether or not a recession is on the immediate horizon, or how severe it will be and how long it will last, if it actually occurs. Nevertheless, a strong case can be made for reauthorization of a contingency fund that would provide some federal assistance to states that face financial burdens from rising caseloads during that time period. Furthermore, the triggers that would enable states to access these funds should be updated in light of developments that have occurred during the past five years; and the level of funds should be uncapped, to ensure that they are sufficient to meet the needs of states.

REFERENCES

Bell, Stephen. "Why Have Welfare Caseloads Fallen?" Discussion Paper, Urban Institute, 2001.

Falk, Gene. "Welfare Reform Financing Issues: Recession Funding." Congressional Research Service Report to Congress, 2001.

Holzer, Harry J. "Unemployment Insurance and Welfare Recipients: What Happens When the Recession Comes?" Policy Brief, Assessing the New Federalism, The Urban Institute, 2000.

Lazere, Ed. "Unspent TANF Funds at the End of Fiscal Year 2000." Center on Budget and Policy Priorities, 2001.

Primus, Wendell and Ed Lazere. "TANF Supplemental Grants Should be Extended for Fiscal Year 2002." Center on Budget and Policy Priorities, 2001.

1. These studies are reviewed and critiqued in Bell (2001).