Evolution of Expenditures and Participation in Public Assistance Programs

For much of U.S. history, public assistance was the responsibility of local townships and counties, with states becoming more and more involved over the last half of the nineteenth and the early twentieth century. As in many other areas of social policy, the federal government's involvement did not begin until the New Deal programs of the 1930s established a precedent for federal responsibility in this area. Even so, such programs remained relatively small in the immediate decades after the 1930s, with only small numbers of recipients and small costs.

Things changed dramatically in the 1960s and 1970s. The primary cash assistance program, Aid to Families with Dependent Children (AFDC), increased dramatically due to a variety of changes that brought many more eligible single mothers and their children onto the assistance rolls.¹ The establishment of the Food Stamp program in the early 1960s resulted in major spending increases during the 1970s, when Food Stamps were expanded to every county and payment rules were simplified. The Medicaid program was established in 1965, providing health insurance to low-income families who met certain eligibility criteria. Medicaid spending levels increased steadily through the 1970s and 1980s, reflecting both increases in the eligible population and increases in medical costs. Highly variable state programs for the elderly and the disabled were moved to the federal level in the early 1970s when the Supplemental Security Income (SSI) program was created to provide uniform cash assistance to elderly, blind, and disabled individuals throughout the nation. Finally, as a supplement to low-wage workers, the Earned Income Tax Credit (EITC) program began on a small scale in 1975.²

Figure 3-1 shows how inflation-adjusted expenditures on public assistance programs have changed since 1965. (Medicaid is not shown in Figure 3-1, but is discussed below.) After peaking in the mid-1970s, AFDC expenditures have been largely constant. Food Stamp expenditures expanded with program expansions in the 1970s, fell during most of the 1980s, but have grown again over the past 7 years as caseloads have grown. The SSI program remained a relatively constant-expenditure program for the first 10 years of its existence, but its expenditures have recently shot upward with expanded eligibility categories. Similarly, the EITC was a small program for its first 10 years, but over the past 10 years, major benefit expansions have made the EITC program as large as AFDC, Food Stamps, or SSI.³

Figure 3-1

Dollars spent on income support (1995 dollars). NOTES: AFDC and SSI show benefit payments. Food Stamps shows coupon value. EITC shows total costs including refunds and tax expenditures. SOURCES: AFDC 1965–1993, Food Stamps 1965–1994, and (more...)

Figure 3-2 shows changes in the number of participants in income support programs. Participation in AFDC was reasonably flat throughout the 1970s and 1980s. Food Stamp participation trends mirror spending trends. In both AFDC and Food Stamps, sharp caseload increases occurred in the early 1990s. As Figure 3-2 indicates, these increases leveled off by the mid-1990s, and more recently available data indicate that caseloads have fallen substantially since 1995 in both programs. EITC participation has risen along with benefit levels. SSI participation has risen only slowly over the past decade, although costs are rising more steeply.

Figure 3-2

Participants in income support programs. NOTES: AFDC, Food Stamps, and SSI show numbers of recipients. EITC shows number of tax returns. SOURCES: AFDC 1965–1993, Food Stamps 1965–1994, and SSI 1994 from Social Security Administration (1995); (more...)

Figure 3-3 provides a direct comparison of inflation-adjusted benefit cost per participant in each of these programs. Per-person AFDC spending declined over time, and per-person Food Stamp spending has been largely flat. The SSI and EITC programs show increases in benefits paid per participant over time. It is also noticeable that SSI recipients receive far more assistance than participants in other programs.

Figure 3-3

Average monthly benefits (1995 dollars). NOTES: AFDC, Food Stamps, and SSI show benefits per person. EITC shows benefits per tax return. SOURCES: AFDC 1965–1993, Food Stamps 1965–1994, and SSI 1994 from Social Security Administration (1995); (more...)

In stepping away from the specific numerical trends, what are the implications of these changes over the past few decades? Cash support for nonelderly and nondisabled individuals has always been relatively limited in the United States, compared to most European nations. As a share of public assistance support, cash support has steadily declined over the past two decades. Increasingly, resources are available through in-kind programs (such as Food Stamps or Medicaid) or through behaviorally tied support programs, in which cash assistance is linked with work behavior. This is most obvious in the growth of the EITC, which is available only to families with working low-income adults. But even the AFDC program became increasingly behaviorally linked, as legislative changes over the past 15 years mandated that more AFDC recipients participate in job search and employment programs.

The United States continues to distinguish sharply between different groups of low-income families. Elderly individuals receive far more support than families with working-age adults. Nonelderly low-income families have always been a source of frustration for the public assistance system. On the one hand, the adults in these families are viewed with some suspicion: why are they not successfully working their way out of poverty? While always applied to male-headed households, this viewpoint has also come to dominate our image of female-headed households as well, as employment among mothers has become more accepted. On the other hand, the adults in these families come attached to children, whom we view with less suspicion and want to assist. The result is a constant tension in public assistance programs between the type of requirements and limits put on assistance to families and the needs of the children in those families. Advocates of greater behavioral requirements and more limited assistance inevitably point to the adults and claim that they need to take more responsibility for their own economic well-being. Opponents of these changes inevitably point to the children and claim that they should not be hurt because of the misfortunes of their parents. Recent legislative changes have supported stronger work mandates, but this debate is far from resolved. Credible research that shows how children are affected by full-time work requirements imposed on their single-parent mothers may have a major effect on future changes in the structure of public assistance programs.

Finally, it is worth noting that the expenditure trends in these four public assistance programs are dwarfed by the growing expenditures in the Medicaid program. Figure 3-4 plots inflation-adjusted spending on Medicaid, separately showing spending on the elderly and disabled (largely those eligible for the SSI program) and spending on families with children (largely those eligible for the AFDC program.) While health expenditures rose throughout the economy, they rose faster for Medicaid, in part because Medicaid generally serves a population with greater health problems.⁴ In recent years, Medicaid spending on the non-elderly, nondisabled population has leveled off, but it has continued to increase for the elderly and disabled. Some of the biggest expenditure increases have been for the Medicaid population in long-term care facilities.

Figure 3-4

Dollars spent on Medicaid by eligibility category (1995 dollars). SOURCES: 1972 and 1994 from Social Security Administration (1995); 1973 from U.S. Department of Health, Education, and Welfare (1975); 1974 from U.S. Department of Health, Education, and (more...)

Trends in the number of Medicaid recipients look quite different from spending trends, as Figure 3-5 indicates. Until the late 1980s, the population of recipients was quite flat, despite steady increases in expenditures. Recent eligibility expansions for children in low-income families have greatly increased the number of young Medicaid recipients. ⁵ Slight increases in recipiency among the elderly and disabled have also occurred. Figure 3-6 shows the implications for Medicaid spending per recipient. Most strikingly, per-person Medicaid expenses for low-income children and related adults have been essentially flat—and are very low compared to per-person expenses for the elderly or disabled. All of the growth in Medicaid dollars for families and children is due to increases in the eligible population. In sharp contrast, per-person Medicaid spending for the elderly and disabled has increased steadily for over 2 decades. Medicaid spending on the elderly and disabled has been largely driven by increases in the person cost of services provided to this population, and not by population growth.

Figure 3-5

Number of Medicaid recipients by eligibility category. SOURCES: 1972 and 1994 from Social Security Administration (1995); 1973 from U.S. Department of Health, Education, and Welfare (1975); 1974 from U.S. Department of Health, Education, and Welfare (1976); (more...)

Figure 3-6

Medicaid spending per person by eligibility category (1995 dollars). SOURCES: 1972 and 1994 from Social Security Administration (1995); 1973 from U.S. Department of Health, Education, and Welfare (1975); 1974 from U.S. Department of Health, Education, (more...)

Public Assistance Programs and Government Spending

Ongoing federal budget deficits have resulted in increased pressure to cut all forms of federal spending. Many states operate under balanced budget requirements and are also constantly seeking areas where costs can be reduced in order to meet other public demands. Public assistance programs have often been a primary target in efforts to cut state and federal budgets. At least some of this is due to a misperception on the part of many Americans about the role played by public assistance in the budget. For instance, a 1995 CBS News/New York Times poll indicated that over 50 percent of the population thinks the federal government spends more than 20 percent of its budget on welfare programs (Roper Center for Public Opinion Research, 1995).

Figure 3-7 shows the composition of federal expenditures in 1995. Public assistance accounted for 14 percent of federal expenditures, of which 6 percent was due to Medicaid spending. While not an insignificant share, more was spent on Social Security, on defense, and on net interest on the debt.

Figure 3-7

Federal expenditures for fiscal year 1995. SOURCE: U.S. Office of Management and Budget (1995).

Much of the growth in federal expenditures on public assistance is relatively recent, and it is heavily due to increases in Medicaid expenditures. Figure 3-8 shows the trends over time on government spending on social programs as a fraction of all outlays. Family support programs (primarily AFDC) have been flat at about 1–2 percent of the U.S. budget for decades. If Food Stamps are added to family support, this accounts for a flat 2–3 percent of the U.S. budget. In these core public assistance programs—the programs that U.S. citizens are most likely to identify as ''welfare"—there is no evidence of high or growing budget shares.

Figure 3-8

Government spending on social programs as a fraction of outlays. NOTES: The category "Family Support" includes payments to states for AFDC benefits, administration, and child support enforcement. The category "All other antipoverty" includes child nutrition (more...)

The federal budget share for all antipoverty programs (except Medicaid) has risen from 6 to 8 percent over the past 5 years, primarily because of the growth in the EITC and SSI programs, discussed earlier. But the big budget-buster is Medicaid, which has increased from 5 to 8 percent of the federal budget in only a few years.

The effect of Medicaid on public budgets is even more visible at the state level. Figure 3-9 shows the breakdown of state expenditures for 1992, the most recent year for which these data are available. Medicaid accounts for 11 percent of the average state budget in that year; other public assistance accounts for only 3 percent. As at the federal level, the Medicaid share of state budgets has been growing dramatically over time. The result is that virtually all states are facing a crisis in their public assistance spending: All states are currently spending more for public assistance in total than they were 10 years ago, but almost all states are spending less on non-Medicaid assistance. In short, programs for low-income nonelderly and nondisabled families have been cut in order to accommodate the growth in spending on medical assistance. While many have decribed the growing cost of public assistance faced by the states, few state governors have publicly discussed the primary cause of this problem—increasing costs in medical services (especially long-term care services) for elderly and disabled persons. Instead, they have often mistakenly assumed that the problem is spending in other programs and have cut general assistance, lowered AFDC benefits, or taken other steps that limit public assistance spending in areas where costs have not been rising.

Figure 3-9

State expenditures for fiscal year 1992. SOURCE: U.S. Bureau of the Census (1993).

The 1996 Legislative Changes⁶

The 1996 welfare reform legislation has been described as a revolutionary change in the structure and emphasis of U.S. welfare programs. In reality, the changes are at once both less radical and more radical than is often claimed. They are less radical in the sense that the criticisms of existing antipoverty programs that they embody are not new and reflect concerns that have long been part of the U.S. debate over helping the poor. Encouraging work, strengthening families, and reducing government costs are not new ideas. On the other hand, this legislation has produced a more fundamental change in the federal government's role in antipoverty efforts than any legislation since the Social Security Act of 1935, giving the states much more control over programs and the federal government much less. This section summarizes some of the most important aspects of the new legislation.

The legislation creates a new block grant to the states, the Temporary Assistance for Needy Families (TANF) block grant. The AFDC program is abolished and states are given almost complete control over the design of their public assistance programs. States can use TANF money for any programs that accomplish the purposes of the block grant, which include providing assistance to needy families, ending the dependency of needy parents on government benefits, preventing and reducing out-of-wedlock pregnancies, and encouraging the formation of two-parent families.

In giving states this expanded program authority, the legislation eliminates family entitlements to cash assistance. Under AFDC, any family that met the eligibility requirements had to be provided with assistance according to established rules and regulations. Under the new law, states have much more discretion in determining who should get funds and how much they should get. They can eliminate some groups from assistance, redirect money away from cash support toward services designed to prevent teen pregnancy and promote marriage, or impose behavioral requirements on the recipients of public assistance funds. If money runs short at the end of a budget year, families can be turned away.

State entitlements to open-ended federal payments are also at an end. AFDC was a matching grant program rather than a block grant program. If states spent more, federal dollars automatically increased. Under TANF, states will receive a fixed amount of money from the federal government in future years, equal to the federal payments they received in the early 1990s for AFDC and related welfare-to-work programs. This leaves states bearing the financial risk should there be an increase in poverty or unemployment; the federal government will not automatically increase payments if a state has to provide assistance to more people. The federal dollars are also fixed at the same nominal level, so they become less over time with inflation.

Although states have greater discretion to determine who is helped, the federal legislation imposes new mandates with regard to work requirements and payment limits for those who receive TANF funds. Any parent who has received 24 months of assistance in programs funded through TANF must be working or in a work program in order to receive further funding. By 1997, 25 percent of all families in the state receiving TANF support must be working at least 20 hours per week. By 2002, 50 percent of all families must be working at least 30 hours per week. (States have substantial discretion to define what counts as "work.") This vastly increases the share of the caseload who must be working, although these requirements are lowered if caseloads fall within the state. (Because of the recent decline in caseloads, in the short run a number of states will face much lower requirements.) The 1996 legislation provided no new federal funds to assist states in expanding their work programs, although some additional funds were added in the 1997 budget legislation.

TANF dollars are time-limited to any individual. No family can receive funding from TANF if an adult in that family has already received 60 months of assistance over his or her lifetime. (At their option, states can impose even shorter time limits.) States are allowed to exempt 20 percent of their caseload from this 5-year limit. States can also continue to support families with state-only funds, which will probably lead to some creative accounting with regard to which families are being supported on federal versus state dollars.

The enactment of time limits is the most dramatically new part of the legislation. At this point, however, it is unclear exactly what the impact of these time limits will be. If there is high unemployment or if many adults hit these time limits at a point when they are clearly unable to hold a full-time job, states may well try to negotiate for more extensive exemptions. In addition, enforcing this lifetime limit effectively will require state and national tracking systems that identify the cumulative months of support received by any individual in any state. But the legislation included no funds for establishing such systems, which may allow some recipients to avoid the time limits in the near term.

In terms of both dollars and numbers of people affected, the biggest impact of the 1996 legislation will be on those who were once AFDC participants and who now are subject to the state-designed programs funded by the TANF block grant. But a number of other provisions in this bill will affect other programs and other groups of people. The legislation sharply limited the availability of public assistance to immigrants (although some of these cuts were restored in later legislation in 1997); limited Food Stamp benefits to nonelderly, nondisabled adults without children; and narrowed eligibility categories for SSI.

Larger Patterns in Program Changes⁷

As noted above, the 1996 legislation pushed further and faster in some directions where changes in public assistance had already been occurring. This section discusses three of the major trends in the structure of public assistance programs over the past decade.

Acknowledgment

Leslie Moscow provided excellent research assistance.

References

• Blank, R. 1997. a. It Takes a Nation: A New Agenda for Fighting Poverty. Princeton, N.J.: Princeton University Press.
• 1997. b. Policy watch: The 1996 welfare reform. Journal of Economic Perspectives 11(1):169–177.
• 1997. c. What Causes Public Assistance Caseloads to Grow? National Bureau of Economic Research Working Paper 6343. Cambridge, MA: NBER.
• Internal Revenue Service Various Statistics of Income—19xx Individual Income Tax Returns. Washington, D.C.
• years (a) Various Statistics of Income Bulletin; Season 19xx. Washington, D.C. years (b)
• Peterson, P. 1995. The Price of Federalism. Washington, D.C.: The Brookings Institute.
• Roper Center for Public Opinion Research 1995. CBS News/New York Time poll conducted April 1–4, 1995. Accessed via Dialogue, file 468, Public Opinion on Line.
• Social Security Administration 1995. Social Security Bulletin, Annual Statistical Supplement: 1995. Washington, D.C.: U.S. Government Printing Office.
• U.S. Bureau of the Census 1993. State Government Finance, 1992. GF/92-3. Washington, D.C.: U.S. Government Printing Office.
• U.S. Department of Health and Human Services 1995. a. Health Care Financing Review: Statistical Supplement. Baltimore, Md.: Health Care Financing Administration, Office of Research and Demonstrations.
• 1995. b. Overview of the AFDC Program: Fiscal Year 1994. Washington, D.C.: U.S. Government Printing Office.
• U.S. Department of Health, Education, and Welfare 1975. Numbers of Recipients and Amounts of Payments Under Medicaid, Fiscal Year 1973. Washington, D.C.
• 1976. Numbers of Recipients and Amounts of Payments Under Medicaid, Fiscal Year 1974. Washington, D.C.
• U.S. House of Representatives 1994. 1994 Green Book: Overview of Entitlement Programs. Washington, D.C.: U.S. Government Printing Office.
• U.S. Office of Management and Budget 1995. Budget of the United States Government, Historical Tables, 1996. Washington, D.C.: U.S. Government Printing Office.

Footnotes

1

As discussed later, AFDC was abolished in the 1996 legislation.

2

These are the programs discussed in this chapter. A host of other programs can also be considered part of the public assistance system, but these are much smaller in terms of both expenditures and participation.

3

Due to recent legislative changes in Food Stamps, AFDC, and SSI, by 1997 the EITC is expected to be the largest of these four programs.

4

Figure 3-4 adjusts Medicaid dollars by the gross domestic product price deflator. Even if they are adjusted by the consumer price index for medical care, Medicaid spending still doubles between 1980 and 1995.

5

In particular, Medicaid eligibility for children was de-linked from family AFDC eligibility. Currently, all children in families below 135 percent of the poverty line are eligible for Medicaid.

6

The description in this section closely follows that in Blank (1997b).

7

This section closely follows parts of the discussion in Chapter 3 of Blank (1997a).

8

Nixon's negative income tax proposal was the high-water mark of efforts to provide large-scale cash assistance to the poor, rather than behaviorally tied assistance. The earlier history of public aid shows an ongoing focus on behavioral regulations as a precondition for assistance.

9

See Peterson (1995) for a discussion of this issue.

10

See Blank (1997c) for a discussion of the causes behind recent caseload changes.