California Work Opportunity and Responsibility to Kids (5180)

In response to federal welfare reform legislation, the Legislature created the California Work Opportunity and Responsibility to Kids (CalWORKs) program, enacted by Chapter 270, Statutes of 1997 (AB 1542, Ducheny, Ashburn, Thompson, and Maddy). Like its predecessor, Aid to Families with Dependent Children (AFDC), the new program provides cash grants and welfare-to-work services to families whose incomes are not adequate to meet their basic needs. A family is eligible for the one-parent component of the program if it includes a child who is financially needy due to the death, incapacity, or continued absence of one or both parents. A family is eligible for the two-parent component if it includes a child who is financially needy due to the unemployment of one or both parents.

The budget proposes an appropriation of $5.1 billion ($1.9 billion General Fund, $153 million county funds, $40 million from the Employment Training Fund, and $2.9 billion federal funds), to the Department of Social Services (DSS) for the CalWORKs program in 2005-06. In total funds, this is a decrease of $521 million, or 11 percent, compared to estimated spending of $5.6 billion in 2004-05. This decrease is primarily attributable to savings from (1) a proposed 6.5 percent maximum grant reduction, (2) a proposed reduction in the amount of income which may be disregarded for purposes of determining a working family's grant, (3) savings from adults reaching their 60 month CalWORKS time limit, and (4) savings from proposed child care reforms.

We note that Congress extended funding for the Temporary Assistance for Needy Families (TANF) block grant through March 31, 2005. The Governor's budget assumes TANF funding will eventually be extended or reauthorized at current funding levels ($3.7 billion annually for California) at least through 2005-06.

Maintenance-of-Effort and Transfers Outside of CalWORKs

The proposed budget achieves General Fund savings of $443 million by recognizing additional countable maintenance-of-effort (MOE) spending on State Department of Education (SDE) Child Care and by increasing the amount of Temporary Assistance for Needy Families (TANF) federal funds used to offset General Fund costs in other programs. We review the recent history of CalWORKs MOE expenditures and TANF expenditures for other programs and comment on the Governor's proposals.

TANF MOE Requirement. To receive the federal TANF block grant, states must meet an MOE requirement that state spending on assistance for needy families be at least 75 percent of the federal fiscal year (FFY) 1994 level, which is $2.7 billion for California. (The requirement increases to 80 percent if the state fails to comply with federal work participation requirements.) Countable MOE expenditures include those made on behalf of CalWORKs recipients as well as for families who are eligible for CalWORKs but are not receiving cash assistance. Although the MOE requirement is primarily met through state and county spending on CalWORKs and other programs administered by DSS, state spending in other departments is also counted toward satisfying the requirement. The 2004-05 Budget Act includes $321 million in countable MOE expenditures outside of the CalWORKs program ($28 million from other DSS programs and $293 million from other departments).

The recognition of additional state-countable expenditures outside of CalWORKs toward the MOE has two impacts. First, it reduces the total amount of state funding that needs to be spent specifically on CalWORKs to meet the MOE requirement. Second, it enables the state to achieve General Fund savings while maintaining compliance with the federal MOE requirements by reducing the General Fund appropriation to DSS for CalWORKs. However, to obtain the General Fund savings, program reductions in CalWORKs are necessary unless there are sufficient federal TANF funds available to cover CalWORKs program costs pursuant to current law.

TANF Expenditures Offsetting General Funds Costs. Federal law permits the expenditure of TANF funds on a variety of programs and activities. It may be expended on any program designed to (1) provide assistance to needy families and children; (2) end the dependence of needy parents on government benefits by promoting job preparation, work, and marriage; (3) prevent and reduce the incidence of out-of-wedlock pregnancies; and (4) encourage the formation and maintenance of two-parent families. Moreover, TANF funds can be spent for any purpose permitted under the AFDC program (the predecessor of TANF) or under AFDC Emergency Assistance (EA). (For example, AFDC-EA could be used for juvenile probation.) Finally, up to 10 percent of TANF funds may be transferred to the Title XX Social Services Block Grant and then expended in accordance with the federal rules pertaining to Title XX.

The 2004-05 Budget Act includes TANF appropriations of $138 million for Child Welfare Services, $51 million for emergency assistance Foster Care, $67 million for youth probation, and $315 million to fund Stage 2 child care costs at SDE. In addition, the act transfers $63 million into the Title XX Social Services Block Grant to fund child welfare services and child care.

Historical Fiscal Policy. In the late 1990s, California had relatively large TANF fund balances in comparison to CalWORKs program costs. This was due to CalWORKs caseload decline (which reduced grant costs) and the relatively slow ramp-up of the costs for CalWORKs welfare-to-work activities and associated child care. Accordingly, budgets enacted during this period attempted to use TANF to offset costs in other programs and to maximize countable MOE expenditures in other departments. This provided substantial General Fund savings while allowing CalWORKs to be fully funded. More recently, the end of the caseload decline, increased costs for employment services and child care, and statutory cost-of-living adjustments (COLAs) have put pressure on TANF and MOE funds. In order for the state to remain at the MOE floor (while continuing to maximize both the counting of MOE funds in other departments and the use of TANF funds to generate savings in other programs) reductions in CalWORKs were necessary. For example, COLAs were delayed and a county incentive program was terminated.

The 2004-05 CalWORKs Budget. The budget act contained two significant changes. First, the Legislature adopted budget control language which limited the amount of SDE child care expenditures that would be counted toward the MOE. Specifically, the language limited countable SDE child care to those expenditures on behalf of families receiving CalWORKs, rather than families who may be eligible for CalWORKs but were not receiving cash assistance. This language reduced the amount of countable MOE expenditures by $153 million. Second, with the sunset of the state Comprehensive Youth Services Act on November 1, 2004, the budget replaced $134 million in TANF funding for local youth probation programs with support from the General Fund in the Board of Corrections. Previously, the state had provided $201 million in TANF funding for youth probation on a full-year basis. These two actions freed up $287 million in TANF ($124 million) and MOE ($153 million) funds to be used for CalWORKs recipients. The actions also resulted in General Fund costs of an identical $287 million.

Governor's Proposal. For 2005-06, the Governor proposes to reverse the decisions made in 2004-05. Specifically, the budget proposes to increase the amount of countable SDE child care expenditures and to fund local youth probation programs with TANF federal funds. In addition, the budget proposes new and expanded TANF transfers to offset General Fund costs in other programs. Figure 1 lists $443 million in new or expanded proposals for creating General Fund savings from the CalWORKs program. In order to free up the TANF funds needed for the proposed transfers shown in Figure 1, the budget proposes grant reductions and other changes discussed later in this chapter. We note that the Legislature rejected similar proposals to transfer TANF funds to Title XX to offset costs in Foster Care and Developmental Services last year.

Legislature Has Flexibility. The proposed TANF transfers in Figure 1 appear feasible, but some require state law changes. Similarly, the proposal to recognize additional SDE child care expenditures to meet the MOE requirement is workable under federal law. In fact, the Legislature could elect to count additional SDE child care expenditures toward the MOE, beyond what the Governor proposes. The proposed budget counts toward the MOE $431 million in SDE child care expenditures on CalWORKs recipients and some CalWORKs eligibles. If the Legislature elected to count all SDE child care spending on CalWORKs eligibles, then total countable SDE MOE spending would be $509 million, an increase of $78 million. However, to obtain $78 million in General Fund savings, additional CalWORKs program reductions would be necessary.

Conclusion. The amount of General Fund support for CalWORKs and the use of TANF funds to offset General Fund costs in other programs are fiscal and policy decisions for the Legislature. Specific decisions regarding countable MOE spending and the amount of TANF transfers are set through the budget act. The Governor has set a goal of $443 million in additional savings from CalWORKs. Based on its overall spending priorities, the Legislature must decide whether it concurs with the $443 million target. The Legislature could seek more or less General Fund savings from the CalWORKs program. After the Legislature sets it goal, the CalWORKs MOE accounting and TANF support for other programs can be adjusted accordingly to fit that target.

In other words, the amount of General Fund support and TANF support for CalWORKs and other eligible programs should be based on the fiscal and policy priorities of the Legislature, not an assumed constraint of remaining below the MOE.

Caseload and Grants

Overestimate of CalWORKs Caseload

We recommend that proposed spending for CalWORKs grants be reduced by $17.4 million in 2005-06 because the caseload is overstated. (Reduce Item 5180-101-0890 by $17,400,000.)

CalWORKs Caseload Trends. In 1994-95, the average monthly CalWORKs caseload (families) was about 921,000. Since then, the caseload has declined every year, eventually dropping to about 481,000 in 2003-04. Most of the decline occurred by 2001-02. Since October 2002, the caseload has been essentially flat.

Governor's Budget Forecast. For 2004-05, the Governor's budget forecasts a slight increase in the caseload to 495,000 since 2003-04. The increase is primarily attributable to onetime factors including the recent migration of certain refugees and the end of extended unemployment benefits. Absent policy changes, the budget forecasts a modest caseload decline to 486,000 cases in the budget year.

The Governor's budget proposes to reduce grants by 6.5 percent and to reduce the earned income disregard. Both of these proposals have the effect of lowering the income threshold at which working families become ineligible for cash assistance. If these are adopted, the average monthly caseload will drop by an additional 14,000 cases during 2005-06.

LAO Caseload Forecast. The most recent actual caseload data from the first quarter of 2004-05 indicate that the Governor's budget has overstated the number of cases (families) by about 0.6 percent. Moreover, the budget has overstated the number of persons (adults and children in the family comprising the case) by 2.8 percent. We note that there is significant uncertainty regarding the number of persons per case, especially the number of adults who have reached their time limit and are removed from the assistance unit (the children continue to receive their grant). Although total persons is usually the best indicator of grant costs, given the uncertainty, we are basing our estimate of caseload savings on the difference between the budgeted number of cases in comparison to the actual number of cases. Based on the 0.6 percent difference identified above, the budget for CalWORKs grants is overstated by $17.4 million. Accordingly, we recommend reducing the budget for cash assistance for by $17.4 million.

Budget Deletes Statutory COLAs and Reduces Grant Payments

The Governor proposes to reduce grant payments by 6.5 percent and permanently eliminate the statutory cost-of-living adjustments (COLAs) currently required each year. These proposals result in state savings of $355 million in comparison to the requirements of current law. In addition, the Governor proposes trailer bill language to delete the October 2003 COLA in the event that the state loses it appeal of the Guillen court case requiring such payment.

Repealing the State COLA. Current law requires that CalWORKs recipients receive a COLA equal to the percent change in the California Necessities Index(CNI) each July. The Governor proposes trailer bill language to eliminate this requirement. The January budget assumed that the CNI would be 4.6 percent and that deleting the CalWORKs COLA would result in savings of $164 million. Based on final data, the CNI will actually be 4.07 percent. Accordingly, the savings from deleting the COLA are reduced to $143 million, based on the actual CNI.

Grant Reduction. The Governor proposes to reduce grants by 6.5 percent. This results in state savings of $212 million. Lowering the maximum grant payment level has the effect of lowering the income threshold at which working families become income-ineligible for cash assistance. As a result, about 7,000 families currently receiving relatively small grants would no longer be eligible for CalWORKs under this proposal.

October 2003 COLA Litigation. As discussed in our Analysis of the 2004-05 Budget Bill (page C-223), the state has not provided the October 2003 COLA. In the Guillen court case, advocates for the state's CalWORKs recipients successfully argued in superior court that the state should provide the October COLA. Currently, the administration is appealing this ruling and an appellate court decision is expected sometime during the second half of 2005. The administration is proposing a trailer bill to delete the October 2003 COLA in the event that the state loses its appeal. This trailer bill would result in avoided costs of $131 million in 2005-06. In the event that Guillen is affirmed by the appeals court, the state faces a budget risk of $222 million in retroactive payment to CalWORKs recipients.

Impact on Recipients. Figure 2 shows the maximum CalWORKs grant and food stamps benefits for a family of three under current law, and what the maximum grant and benefits would be under the Governor's reduction proposals. In general, food stamps increases offset about 45 percent of the proposed grant reductions. As the figure shows, under the Governor's proposals, in 2005-06 the maximum monthly CalWORKs grant for a family of three in a high-cost county would be $676, compared to $778 under current law. The maximum monthly CalWORKs grant for a family of three in a low-cost county would be $644 under the Governor's proposals, compared to $742 under current law. (For purposes of discussion, Figure 2 assumes that restoration of the October 2003 COLA is part of current law.)

As a point of reference, the federal poverty guideline for 2004 (the latest reported figure) for a family of three is $1,306 per month. (Federal poverty guidelines are adjusted annually for inflation.) Under current law, the combined maximum CalWORKs grant and food stamps benefits in high-cost counties would be $1,057 per month (81 percent of the poverty guideline). Under the Governor's proposals, combined benefits in high-cost counties would instead be $1,001 per month (77 percent of the poverty guideline). Combined benefits in low-cost counties would be $1,037 per month (79 percent of the poverty guideline) under current law, compared to $983 (75 percent of the poverty guideline) under the Governor's proposals.

Other Budget Issues

Unspent TANF Funds Identified

The Governor's budget has identified $407 million in additional unspent Temporary Assistance to Needy Families (TANF) funds from past years in comparison to the May 2004 estimate. We review the sources of these additional TANF carry-over funds. To reduce future overbudgeting, we recommend eliminating the child care "hold back" and midyear supplemental allocations to counties.

Background. Each year California receives from the federal government its $3.7 billion annual TANF block grant. Any unspent TANF funds are carried forward on California's books for expenditure in future years. The 2004-05 Budget Act assumed that $115 million in federal TANF funds would be carried forward from 2003-04 and be available for expenditure in 2004-05.

Additional Carry-Forward TANF Funds Identified for 2004-05. The Governor's budget has identified $407 million in additional TANF carry-forward funds as compared to the May 2004 estimate. Figure 3 shows the sources for these additional funds. Specifically, about $182 million is from an accounting error pertaining to 2001, and $202 million is from unspent county allocations from 2002-03 and 2003-04. (Under current law counties receive one block grant to cover the estimated costs of administration, welfare-to-work services [including case management], and child care.) This block grant is known as the single allocation, and counties are free to move funds among the different program components funded in the block grant. Under current practice, DSS increases block grants midyear if it identifies additional costs pertaining to caseload, workload, or court cases.

History of Finding Additional Carry-Forward Funds. The identification of additional carry-forward funds following enactment of a budget is not an isolated incident (though the magnitude for this year is unprecedented). For the past four fiscal years, the January budget has identified an average of $111 million in additional carry-forward TANF funds in comparison to budgets enacted about six months earlier.

What Causes Unspent Funds? As noted above, a major factor contributing to unexpended carry-over balances was the counties' inability to expend their single allocation block grant funds. This amounted to $143 million in overbudgeting for 2003-04. We believe that most of the unspent funds are related to uncertainty regarding the amount of the single allocation, and cautious budgeting practices by counties which limit expenditures until final allocations are known. Below we discuss three possible causes of uncertainty and unspent funds.

• Counties Receive Midyear Increases Too Late to Enable Their Expenditure. Counties receive notification of their initial planning allocation just before the start of the fiscal year. Soon after the state budget is enacted, a final allocation letter is sent to the counties. For 2003-04, the final allocation was sent in September 2003. At the time the Governor's budget is released each January, supplemental allocations are discussed with counties and a supplemental allocation letter (noting that additional funds are contingent on legislative approval) is sent to the counties within a few weeks of the release of the budget. During 2003-04, counties were notified of a supplemental allocation of $48 million in February 2004. On April 22, counties received final notification that the February funds were approved. Finally, counties were notified of an additional $9 million allocation on June 30, 2004, the last day of the fiscal year.
• The Child Care Hold Back. The state holds back approximately 5 percent of estimated costs for Stage 1 and Stage 2 CalWORKs child care in a reserve (about $60 million). The reason for the hold back is that Stage 2 is controlled by SDE while Stage 1 is controlled by DSS, and there is some uncertainty as to how many families will move from Stage 1 to Stage 2 each year. Under this system, counties do not know how much of these hold back funds they will eventually receive until notified by the administration well into the fiscal year. Thus, counties may tend to reduce spending on program components they directly control such as eligibility administration and employment services, so that they can fund child care (which they do control because it is virtually an entitlement). This way, counties will be sure they have enough funds to cover program expenses, even if they do not receive the child care reserve funds they have requested.
• General County Fiscal Concerns. During 2003-04, there was substantial uncertainty about the payment of vehicle license fee subventions to the counties. As a result, many counties adopted county-wide hiring freezes. Since most single allocation funds are for county staff, these freezes could have contributed to unspent funds.

Issues for Legislative Consideration. Overbudgeting means that TANF funds are held at the county level and are not available for other program needs such as grant costs, or for transfer to other programs for the purpose of creating General Fund savings. To avoid future overbudgeting, we discuss some advantages and disadvantages of two possible changes.

• Eliminate the Child Care Holdback. Under this option, all estimated child care funds would be allocated to Stage 1 and Stage 2 with no holdback, thereby reducing uncertainty. In the event the allocation between Stage 1 and Stage 2 is incorrect, both counties and SDE's alternative payment providers usually have unspent funds to cover deficiencies. Thus, recipients are not likely to be impacted by this change.
• End Midyear County Block Grant Supplemental Allocations. Under this approach, county allocations would not be increased midway through the current year for cost increases. However, the budget year proposed allocation would reflect any such costs pressures. This approach would free up TANF funds for expenditure by the Legislature in the budget year instead of maintaining them in county allocations which would not revert for another 15 months. This would give the Legislature more control over scarce TANF funds that could be used for priorities both inside and outside of the CalWORKs program. The downside of this approach is that supplemental allocations are made in order to address unanticipated costs, usually related to caseload. Without midyear supplements, some counties may have difficulty serving all clients with the necessary case management and employment services.

Analyst's Recommendation. In order to increase legislative authority over TANF funds, we recommend adopting the changes presented above. We also recommend working with stakeholders to develop a "county request" program, for counties who can demonstrate both midyear cost increases related to workload and the ability to expend the additional funds.

Current-Year Costs Overstated

Our review of actual caseload and expenditure data through October 2004 indicates that the Governor's budget overstates the CalWORKs costs for 2004-05 by $118.5 million. We recommend that the Legislature recognize these savings and increase the Temporary Assistance to Needy Families carry-forward balance available for 2005-06 by $118.5 million. (Increase Item 5180-101-0890 by $118.5 million.)

Our review indicates that DSS has over-estimated program costs for 2004-05 for both cash grants and county single allocation expenditures. We discuss each overestimate below.

DSS Budgeting Process for Cash Grants. To estimate the budget for cash assistance payments in the current year, DSS determines the cost per person based on the actual trends from prior years and then applies the cost per person to its latest estimate of the future caseload. The department then adjusts the projected budget based on recently enacted policy changes that would not be captured in recent expenditure trends (for example, the estimated impact of the recently adopted welfare reform measures which were not implemented until December 2004). At the time DSS prepared the budget, it had access to expenditure and caseload trends through July of 2004.

More Recent Data Available. Subsequently, we reviewed caseload data through September 2004 and cash expenditure data through October 2004. We then applied the cost per case in September 2004, to the department's estimated caseload from November 2004 through June 2005 to arrive at a baseline budget for 2004-05 cash assistance. We then adjusted this baseline to reflect policy changes not captured in the trend through September. We conclude that DSS has overestimated grant costs for 2004-05 by $96.5 million.

Budgeting Practice for County Block Grants. With respect to 2004-05 county block grant allocations for CalWORKs administration, welfare-to-work services, and Stage 1 child care, DSS assumes that counties will spend their entire allocation. In addition, DSS assumes that counties will also expend an additional $22 million that will be allocated to them in a supplemental letter released in February 2005.

History of Unspent County Allocations. As discussed in the previous issue, counties have a history of not spending all of their block grant funds. One source of these unspent funds is late notification of additional county block grant funds. Based on past history, we conclude that, at a minimum, counties are not likely to expend the most recently allocated $22 million in block grant funds. Thus, the budget overstates county expenditures by $22 million and underestimates the likely TANF carry-forward by an identical amount.

Analyst's Recommendation. As discussed above, we estimate that cash grants and county block grant expenditures are overstated by a total of $118.5 million. We recommend that the Legislature recognize these savings by increasing the TANF carry-forward balance for 2005-06 by $118. 5 million. These TANF funds are available for expenditure in the budget year or may be held in reserve for future years. We will monitor actual expenditure and caseload trends over the coming months and will advise the Legislature of any changes in our estimates for 2004-05.

Proposal to Reduce Earned Income Disregard

Under current law, the first $225 of earned income and 50 percent of each additional dollar earned is disregarded (not counted) for purposes of determining a family's grant amount. The Governor proposes to reduce the disregard factors to $200 and 40 percent. This proposal reduces the grants for all working recipients, results in savings of $80 million and probably has minimal impact on the work incentive of CalWORKs recipients. We comment on the Governor's proposal and present alternative approaches which would likely increase the work incentive of CalWORKS recipients, but would result in lower budgetary savings.

Background. The maximum CalWORKs grant is the amount of money a family receives if it has no other income. If the family has income, the grant is reduced after a specified amount of income is not counted (referred to as the income disregard). In order to provide an incentive for CalWORKs recipients to work, current law disregards (does not count) the first $225 in earned income and 50 percent of each additional dollar earned when determining a family's grant amount. In the nearby shaded box, we show how the disregard works in determining the grants for families with two different levels of income.

Exit Point for Cash Assistance. The maximum monthly grant, in combination with the disregard policy, creates the exit point for CalWORKs (the point at which a family is no longer financially eligible for the program). For a family of three in a high-cost county the current exit point is a monthly income of $1,671 (128 percent of the 2004 poverty guideline). When food stamps and the earned income tax credit are added, the family's total income at the exit point is about $1,939 (149 percent of the poverty guideline). Figure 4 shows how grants decline as earned income rises. As the figure shows, the grant for a family of three reaches zero (the exit point) before $1,700 per month.

Governor's Proposal. The governor proposes to reduce the earned income disregard factors to $200 and 40 percent, effective October 1, 2005. This would reduce the grants for all working recipients. This is because reducing the earned income disregard has the practical effect of increasing the amount of income that is counted when determining a family's grant level. The DSS estimates that for cases with income, the average grant reduction will be about $79 per month. Figure 5 compares the combined earned income and CalWORKs grant for a family of three in a high-cost county under both current law and the Governor's proposal at four different income levels. As the figure shows, working families would have less income (combined earnings and grant) under the Governor's proposal as compared to current law. The greatest impact is on families with the greatest income. A family working 40 hours per week at $9 per hour would see their combined grant and income decline by $148 per month, a reduction of over 9 percent.

Impact on Eligibility. Reducing the income disregard lowers the point at which families would no longer be eligible for a grant. As noted above, the exit point is currently $1,671 per month for a family of three. Under the Governor's proposal, the exit point for a similar family would drop to $1,405 per month (108 percent of poverty guideline). Reducing the exit point will mean that about 8,900 families will become ineligible for CalWORKs. Such exiting families, however, would remain eligible for food stamps, child care, and Medi-Cal, so long as their income remains below the eligibility thresholds for these programs.

Estimated Savings. The Governor's budget estimates that reducing the earned income disregard will result in grant savings of $80.4 million, based on nine months of implementation starting in October 2005. In addition, the budget includes administrative savings of $1.5 million due to the case exits noted above. The budget assumes automation costs of $2.5 million for reprogramming welfare automation systems to reflect the new disregard. For 2005-06, net savings from the proposal would be $79.4 million. In 2006-07, the savings would increase to about $109 million, based on a full-year of operation. These savings assume no behavioral response by CalWORKs recipients. In other words, these estimates do not assume either an increase or decrease in the amount of work performed by CalWORKs families.

Impacts on the Work Incentive and Behavior. Reducing the disregard could have two impacts on the working behavior of recipients. On the one hand, it could result in a disincentive to work by reducing the amount of income retained from starting work or increasing one's hours of work. Thus, new entrants to CalWORKs who have no income along with currently aided families who are not working would be most affected by this disincentive. On the other hand, reducing the disregard could increase the incentive to work to the extent that families decide to work more hours in order to make up for the grant reduction pursuant to the revised disregard. Families with substantial earnings, but with incomes below the exit point would be most affected by this increase in the incentive. With more hours of work, such a family could make up for the lost income and possibly leave cash assistance. We would expect these two effects to in part offset each other.

Given the relatively small change in the disregard structure proposed by the Governor, we would expect minimal net change in the propensity of CalWORKs families to work. Accordingly, the Governor's savings estimates (which assume no change in the work behavior) are reasonable.

Alternatives to the Governor's Proposal

The earned income disregard is a key component of the CalWORKs program. It is a significant part of the work incentive for CalWORKs recipients and, in conjunction with the maximum grant, sets the exit point for the program. However, other factors such as the earned income tax credit and food stamps benefits also affect family income and the incentive to work. Before examining alternative approaches to the earned income disregard, we compare California's disregard to the disregard policies in ten other large states.

Comparison of Disregards to Other Large States. Figure 6 presents the earned income disregard and the exit point for cash assistance for the ten largest states in 2003 (the most recent available data). As the figure shows, California's disregard is among the most generous (only Ohio disregards more earned income), and California has the highest exit point for cash assistance. As a point of reference, Figure 6 also shows the exit point relative to the 2003 poverty guideline.

Comparison of Participation Rates to Other Large States. The earned income disregard is intended as an incentive that affects work participation. Figure 7 compares the disregard in each state to the percentage of families meeting federal work participation requirements through just employment, or through employment and other activities. The states are ranked in order of the percentage employed. In general, states with more generous disregards have higher levels of employment among their caseloads. For example, of the five states with the more generous disregards, four (Illinois, New York, California, and Ohio) have the highest employment rates among their program participants. There are, however, important exceptions. Michigan and New York, which clearly have less generous disregards than California, also have higher participation than California. Texas, with the least generous disregard, still manages the sixth best participation rates among these states. Clearly, other factors besides the disregard affect the work behavior of recipients. These factors include local economic conditions, sanction policies, and the ability of recipients to meet participation through activities other than unsubsidized employment.

Because the Governor's proposal has minimal impact on the work incentive, we discuss two alternatives.

Disregard All Income at a Constant Percentage Rate. Given that Illinois had the highest level of work participation among the ten largest states (see Figure 7), we first examined their disregard. In Illinois, 67 percent of all income is disregarded; however, there is no minimum income amount which is disregarded at 100 percent. (For example, California disregards 100 percent of the first $225 in earnings.) Assuming no change in work behavior, adopting a straight 67 percent disregard in California would actually cost more than current law (at least $40 million per year). It would also raise the exit point for CalWORKs by about $170 for a family of three. A variant on this approach would be to set the disregard rate at 57 percent. This would keep California's exit point near where it is today and would result in annual savings of about $65 million compared to current law. (When welfare reform was first debated in 1997, the Wilson administration initially proposed a 54 percent disregard.)

The main advantage of a constant disregard at a rate higher than the current 50 percent is that is creates a stronger incentive to increase earnings as seems to be illustrated by Illinois. For example, a family earning more than $225 per month who increased their monthly earnings by $100 would keep $67 under the Illinois style disregard compared to just $50 under California's current law. The main disadvantage is that families earning less than $225 would have less of an incentive to work since they could only keep 67 percent of their earnings (instead of the entire amount under California's system). Consequently, such families would be worse off financially.

Disregard More Income at Higher Earnings. Under this approach the 50 percent disregard would apply to all families, but the 100 percent exclusion on the first $225 earned would only be provided to families earning $600 or more per month. Those earning less than $600 (about 20 hours per week at the minimum wage), would receive a flat disregard of 50 percent, but they would not receive the base 100 percent disregard on their first $225 in earnings.

The reason for selecting the $600 amount is that it corresponds to roughly 20 hours of work per week at the minimum wage. Under current law, adult participants must meet a "core" participation hour requirement of 20 hours per week. Unsubsidized employment is one way to meet the core requirement.

This approach would result in annual savings of about $48 million compared to current law. The advantage is that it would strongly encourage recipients to work at least 20 hours per week, because they would receive the benefit of the $225 exclusion once their earnings reach $600 per month. The disadvantage is that it would lower the grants for families with earnings below $600 per month because until they earned $600 they would not receive the 100 percent disregard on their first $225 earned. It would also reduce the incentive for those not working to begin work at less than $600. It would not change the exit point for CalWORKs in relation to current law.

Federal Welfare Reform Reauthorization Clouds This Issue. Under current federal law, states must have certain percentages of their families working or participating in program activities. These percentages are reduced by the "Caseload Reduction Credit," which is the amount of caseload reduction that has occurred since the enactment of the federal 1996 welfare reform legislation. Some versions of welfare reform reauthorization continue or modify this caseload reduction credit. Other versions eliminate this credit. All versions increase the percentage requirements for participation. Because welfare reauthorization provisions are unknown, it is difficult to determine which type of disregard polices will be most advantageous to satisfying revised federal work participation rates.

Conclusion. Whether to change the earned income disregard is a policy decision for the Legislature. The Governor's proposal results in budgetary savings of $80 million and would likely have minimal impact on the work incentive. The alternative approaches described above would probably increase the work incentive, especially the incentive to work more than half-time. On the other hand, they result in less budgetary savings and reduce grants for families with the lowest earnings.

County Performance Rewards and Sanctions

Effective in 2006-07, the Governor proposes to increase or decrease county block grant allocations by up to 5 percent based on county performance in meeting specified participation goals during 2005-06. In advance of this bonus/sanction system, the Governor's budget assumes that counties will increase the hours of employment for recipients resulting in grant savings of $22 million during 2005-06. We review and comment on the Governor's proposal.

Governor's Proposal

The Governor proposes to establish two performance measures for counties: first, that counties increase the rate of employment among their recipients, and second that counties increase the percentage of their recipients meeting federal work participation requirements. (Federal law recognizes activities other than employment, such as vocational education, as counting towards the work participation rate.) Beginning in 2006-07, the Governor proposes to hold back 5 percent of each county's single allocation block grant (excluding child care). Based on county performance for each participation measure, counties could lose or gain up to 2.5 percent of their single allocation funds. It is our understanding that DSS will establish different improvement benchmarks for each county based on local economic conditions and existing performance. The budget assumes that counties will change their approach to welfare-to-work services so as to increase employment among recipients during 2005-06, in advance of the bonus/sanction system being implemented in 2006-07. Based on this increase in employment, the budget estimates grant savings of $22 million in 2005-06.

Previous Incentive Program

Prior to 2000-01, the state law provided that CalWORKs savings resulting from (1) exits due to employment, (2) increased earnings, and (3) diverting potential recipients from aid with one-time payments, would be paid to the counties as performance incentives. Chapter 108, Statutes of 2000 (AB 2876, Aroner), changed the treatment of performance incentives in several important ways. Among these changes, it:

• Prohibited counties from earning new incentives beginning in 2000-01 until the estimated prior obligation owed to the counties had been paid by the state.
• Made future performance incentive payments subject to annual budget act appropriations, rather than being treated as an "entitlement."

By the end of 1999-00, the last year for which an appropriation for new performance incentives was made, counties had earned approximately $1.2 billion in incentive funds. However, the state has paid the counties about $900 million of these incentives, leaving a balance of about $300 million.

Current County Performance Measure

Under current law, counties have only one performance measure for which they are financially accountable. Specifically, in the event that the state fails to meet federal work participation rates, counties would share proportionately in a federal penalty. Under federal law, states are subject to specified federal work participation rates. However, states may reduce the required work percentage rates by reducing their caseload through a mechanism called the caseload reduction credit. Because California (like all other states) has substantially reduced its caseload, its caseload reduction credit means that California's required rate of work participation was 6.7 percent as of 2002. Given this very low performance standard, it may be time to consider adding new performance measures for counties.

Problems With Governor's Proposal

Although consideration of how to improve state and county performance with respect to employment levels and federal participation rates is warranted, we have identified two problems with the Governor's approach.

Performance Measures Reflect Federal Rather Than State Goals. The Governor's selected performance measures are more reflective of federal participation activities than state participation activities. Specifically, California allows recipients to participate in certain activities (such as mental health and substance abuse treatment) which are not countable toward the federal work participation rates. If the Legislature adopts new performance measure for counties, such measures should be consistent with state participation activities and priorities.

Holdback Would Be Disruptive. As discussed in a previous issue, counties have had substantial unspent block grant funds. One source of these funds is uncertainty related to the ultimate receipt of child care hold back funds. Holding back additional county funds, as the Governor proposes, would probably compound the problem. Even in counties that perform relatively well on participation and employment, the data needed to verify their performance will not be available until the latter half of the fiscal year. Thus, even higher performing counties would not know their final allocation and would probably need to reduce spending to allow for the possibility of not receiving the held back funds. Moreover, for counties that fail to meet the performance measures, taking resources away in the form of block grant reductions is likely to compound rather than help their performance problems. This is because it may be difficult for counties to achieve the same performance with less resources.

Alternative Approaches to Funding and Incenting County Performance

Funding County Incentives. As discussed above, counties have about $300 million in earned, but unpaid performance incentives. If the Legislature decides to adopt new county performance measures, it could reward high performing counties by paying off the previous incentives. The payments could be limited to a certain amount each year, and could be subject to the identification of additional carryover funds from prior years. For example, counties could be offered payments of up to $40 million in 2006-07 based on performance in 2005-06 subject to the condition that unexpended TANF carryover funds from 2004-05 are identified by 2006-07. This approach avoids the disruption associated with the Governor's proposed hold back of single allocation funds.

Improving County Performance. Another way of increasing the county incentive to move recipients into employment and self-sufficiency would be to increase the counties' share of grant costs. Currently counties pay just 2.5 percent of grant costs. The state pays the remainder with General Fund and TANF federal funds. If counties had a higher share of grants, for example 10 percent, then they would have a substantial incentive to move recipients into employment so as to reduce the county costs for grants. In order to avoid creating a reimbursable state mandate, offsetting county savings may need to be provided. For example the county share for Adoptions Assistance payments could be reduced, a program over which counties have little policy control; so as to offset the increase in CalWORKs costs, a program over which counties have more control from the point of view of offering employment and education-related services.

Conclusion. Given the limited performance measure in current law, the Legislature should consider adding county performance measures. The Governor's proposed performance measures are reflective of federal priorities, not necessarily state priorities. The Legislature should ensure that any new performance measures are reflective of its policy priorities. The proposed hold back is likely to be disruptive, could compound the existing problem of unspent funds, and may leave low performing counties with insufficient funds to improve their performance.