Policy Brief #7

Has Welfare Reform Changed State Expenditure Patterns?
September 2006

Prepared from a paper by Therese J. McGuire , Kellogg School of Management and David F. Merriman , Loyola University.

Download printable version. (PDF)



• States did not disproportionately cut cash welfare spending during the most recent recession despite their new flexibility via block grants to do so.

• Social assistance spending became more counter-cyclical after the 1996 welfare reform while total state spending became more pro-cyclical.

• State spending on public welfare programs increased as a share of total state spending, from 19.4 percent in 1980 to 26.9 percent in 2003. This growth was driven largely by medical vendor payments for Medicaid.

• Medicaid spending has far outpaced other social assistance spending over the past twenty-five years.


The passage of the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) in 1996 changed the way the federal government finances cash welfare to low-income families. Instead of a matching grant, states receive a set amount of federal funds in the form of a block grant with no adjustments for inflation, caseload numbers, or state spending. Social assistance programs, however, tend to experience increased demand in recessions – more people require assistance from the government during difficult economic times. To meet increased demand, states have to increase spending, but recessions usually bring lower revenues to states, requiring some cuts to discretionary spending. Analysts were concerned that PRWORA’s change to a block grant would jeopardize social spending during times of state fiscal stress and cause states to disproportionately cut welfare spending during recession.

In their paper, “State Spending on Social Assistance Programs over the Business Cycle,” which will appear in the National Poverty Center conference volume Working and Poor: How Economic and Policy Changes Are Affecting Low-Wage Workers,1 Therese J. McGuire and David F. Merriman compare states’ spending in response to the most recent recession in 2001—the first since welfare reform in 1996—with spending and revenue patterns in two prior recessions. They find that states did not cut social welfare spending disproportionately during this last recession. In fact, state spending on social welfare programs (cash assistance and Medicaid combined) was more responsive to economic distress than in pre-welfare reform recessions.

Social Assistance Program Funding

Historically, cash assistance (now Temporary Assistance for Needy Families or TANF, previously Aid to Families with Dependent Children or AFDC) and Medicaid comprised the bulk of state social assistance spending. Prior to welfare reform, the federal government provided between 50 and 85 cents in matching funds for every dollar a state spent on both of these statefederal programs.

PRWORA dramatically changed the way the federal government funded cash welfare. Instead of matching funds, states now receive a block grant that is usually equal to the amount of federal aid a state received for its AFDC program in 1994. Medicaid remains a matching grant program. States are required to maintain their cash assistance funding levels at a minimum of 75 percent of their 1994 spending, called a “Maintenance of Effort” (MOE) requirement. States keep any surplus funds each year and have flexibility to use them in future years for any TANF-related programs. With the rapid welfare caseload declines of the 1990s—as work mandates and cash assistance time limits were enacted amid a robust economy—many states accumulated considerable surpluses.


Changes in Social Spending at the State Level

State and local spending for both Medicaid and AFDC increased dramatically in the late 1960s and early 1970s. In the 1980s, however, federal requirements for these two programs changed drastically. State demonstration projects and caseload declines in AFDC led to cuts in funding. At the same time, the federal government was requiring that more people be covered under Medicaid and that a wider variety of vendors, such as hospitals and nursing homes, be paid with Medicaid funds – leading to a very rapid growth in Medicaid spending.

Diversification of the Safety Net. While spending on TANF declined from the early 1990s through 2003, spending on other social programs such as the Earned Income Tax Credit (EITC) and Supplemental Security Income (SSI) rose dramatically (see Figure 1). Spending for food stamps declined for a time but increased again beginning in 2000.2 Overall, spending on all social assistance programs was higher in the post-reform era (1998–2003) with the exception of cash welfare, which fell 27 percent, mainly owing to significant caseload declines. Expenditures follow caseload fluctuations, except for SSI, which saw a significant rise in expenditures but only a small rise in caseload.

Social Assistance Spending as a Proportion of Total State Spending. State spending on public welfare programs3 also increased as a share of total state spending, from 19.4 percent in 1980 to 26.9 percent in 2003. This growth was driven mainly by medical costs. The share of state budgets devoted to medical vendor payments rose from 8.3 in 1980 to 18.1 percent in 2003. Cash assistance, in contrast, declined from 3.3 to 1 percent of state spending. The turnaround in spending is perhaps best illustrated by the shifting amounts of federal aid to states. In 1980, federal aid for income security programs such as AFDC, food stamps, housing assistance, and child nutrition programs were about the same as federal support for Medicaid. Only 23 years later, federal grants for Medicaid were nearly twice the federal grants for income security programs. (See Figure 2) During the same period, there was a 20 percent increase in grants for cash welfare compared to a 372 percent increase for Medicaid.

State Revenues and Expenditures: Responses to Three Recent Recessions

The extent to which economic downturns affect state spending on social programs depends on how revenues and expenditures respond to those downturns. McGuire and Merriman compare state revenues, total state spending, and state spending on education, medical vendor payments, and cash assistance during the three most recent recessions – early 1980s, early 1990s, and 2001, to illuminate how revenues declined and how spending responded over the last three recessions. They use these comparisons to determine whether the first recession after welfare reform differed significantly from previous recent recessions.

Revenue trends. Examining total revenue, revenue from general sales and gross receipts, and individual income tax revenue, the authors find that the major revenue impacts for the 2001 recession occurred after the recession, not during it, as with the previous two recessions. Additionally, in the recessions of the early 1980s and early 1990s, revenue from income taxes rose rapidly as soon as the recessions ended. During 2001, however, revenues did not fall until after the recession began. Revenues then plummeted over the next year and a half with no signs of recovery. This dramatic fall in income tax revenues is the defining characteristic of the state fiscal crisis of the early 2000s. Over all three recessions, the authors did not find any significant difference in federal funding to states for income security programs or Medicaid.

Spending trends. Previous studies conducted by the Urban Institute and the Department of Health and Human Services that examined how state spending responded to the 2001 recession found that states were able to get through budget crunch without major cuts to social welfare spending, including Medicaid, because of TANF surplus funds and the MOE requirement. McGuire and Merriman find, however, that total state spending actually increased by 3.5 percent in the early 1990s recession and by almost 7 per cent in the early 2000s recession with no years of decline in either period despite the decline in state revenues states. This was also the case for two major components of state spending – education and Medicaid. The decrease in cash assistance spending in the most recent recession is likely the result of caseload declines that were not significantly reversed by the 2001 recession.


Has Social Assistance Spending Become More Sensitive to Business Cycles after Welfare Reform?

McGuire and Merriman investigate whether states disproportionately cut or expand social welfare spending during economic downturns. They compare a percentage change in the unemployment rate with each of various categories of state government spending from 1979 to 2003 using data from all 50 states. Their analysis accounts for state differences, national level policy changes, national economic conditions, Republican control of state branches of government, and a variable to assess whether spending became more sensitive to the economic cycle after PRWORA was implemented.


They find no evidence that states disproportionately cut spending on social welfare programs during economic downturns – total state spending declined more in the 2001 recession than in prior recessions, making it more pro-cyclical. Each one percent increase in the unemployment rate was associated with an additional $61 per capita drop in state spending over what would have occurred before PRWORA was implemented. Education spending also appears to have become more pro-cyclical – almost six to seven times more responsive to unemployment after 1997 than before. 

Despite the change from federal matching funds to a block grant, social assistance spending appears to have increased with increases in unemployment – becoming more counter-cyclical after welfare reform. McGuire and Merriman find that a 3 percent increase in the unemployment rate is associated with a 13 percent increase in social assistance spending, or an $87 per capita increase. Nearly 2/3 of this increase is attributable to increases in medical vendor payments. These results may be tentative considering the limited data available since 2001, the relatively low unemployment rates since welfare reform, and the large TANF surpluses states had accumulated going into the most recent recession. Given the current information, however, the authors suggest that states have the ability and resources necessary to maintain social assistance spending during mild economic downturns. 


Policy Implications

That state spending on cash assistance remained counter-cyclical after welfare reform and was even more responsive to economic distress than before welfare reform may be a consequence of the diversification of the safety net. The EITC, Food Stamps and SSI, among other federally-funded programs, have become much more important components of the social safety net than cash welfare. These programs now play a much larger relative role in securing the well-being of lowincome families and individuals.

The explosion in Medicaid spending and the large numbers of low-income and disabled people that depend on this program have become a major concern for states and will continue to be a budget problem even if state revenues stabilize.

End Notes

1 The volume, Working and Poor: How Economic and Policy Changes Are Affecting Low- Wage Workers, edited by Blank, Danziger and Schoeni will be published in December 2006. It can be ordered from the Russell Sage Foundation: http://www.russellsage.org/publications/ books/060711.765292

2 The federal government is responsible for all spending on the EITC, the basic SSI grant, and Food Stamps.

3 Public welfare programs include medical vendor payments, cash assistance and other public welfare programs. The other category includes spending on highways, police protection, corrections, natural resources, parks and recreation, and government administration. 


About the authors

Therese J. McGuire is Professor of Management and Strategy at the Kellogg School of Management and a faculty fellow at the Institute for Policy Research at Northwestern University.

David F. Merriman is Professor of Economics at Loyola University, Chicago.


The National Poverty Center is charged with promoting high-quality research on the causes and consequences of poverty, evaluating and analyzing policies to alleviate poverty, and training the next generation of poverty researchers.

Rebecca M. Blank and Sheldon H. Danziger, Co-Directors

Major funding for the National Poverty Center is provided by the Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services.

Any opinions, findings, conclusions, or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the view of the National Poverty Center or any sponsoring agency.