Generational Differences In U.S. Public Spending, 1980–2000

Abstract

The balance between spending on children and spending on the elderly is important in evaluating the allocation of public welfare spending. We examine trends in public spending on social welfare programs for children and the elderly during 1980–2000. For both groups, social welfare spending as a percentage of gross domestic product changed little, even during the economic expansions of the 1990s. In constant dollars, the gap in per capita social welfare spending between children and the elderly grew 20 percent. Unlike spending for programs for the elderly, spending for children’s programs suffered during recessions. Public discussion about the current imbalance in public spending is needed.

Recent budgetary proposals continue to transfer federal responsibility for social welfare programs to the states. The passage of the 1996 Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA) marked a dramatic shift in social policy by ending standard entitlement to public assistance for low-income children and families and by devolving federal oversight of traditional welfare mechanisms to the states.¹ There are now proposals to turn Medicaid, the single largest insurer of U.S. children, from an entitlement with an open-ended funding commitment from the federal government into a state “block grant” with a predetermined allotment of federal funds that would also allow states to cap enrollment and spending for optional programs within Medicaid.² In contrast, Social Security and Medicare are federally administered through trust funds buffered from short-term economic fluctuations.

In comparison with other countries, social indicators for U.S. elders are quite good, whereas indicators for children are poor, and evidence suggests that social welfare spending can, to some extent, reduce poverty.³ Despite recent declines, U.S. child poverty rates are among the highest in the world.⁴ Over the past twenty years, the absolute number of children living in poverty has been nearly three times the number of elderly living in poverty, and poverty rates for children have consistently exceeded those for the elderly (Exhibit 1).⁵ In spite of these discrepancies, recent and anticipated increases in the number and proportion of U.S. elderly have been accompanied by a “graying” of the federal budget.⁶ For example, the 2004 Medicare trustees’ report projects that Medicare spending will grow from 2.6 percent of gross domestic product (GDP) in 2003 to 7.75 percent in 2035.⁷ In contrast, U.S. social spending devoted to children, including health care, is among the lowest in the world.⁸

EXHIBIT 1.

Percentage Of U.S. Children And Elders Living In Poverty, 1980–2000

SOURCE: U.S. Bureau of the Census, Current Population Survey, Annual Social and Economic Supplements, Poverty and Health Statistics Branch/HHES Division, “Table 3. Poverty Status of People, by Age, Race, and Hispanic Origin: 1959 to 2002,” Historical Poverty Tables, www.census.gov/hhes/poverty/histpov/hstpov3.html (21 June 2004)

In a 1991 study, Ted Benjamin and colleagues found that during the recession of the early 1980s, real spending on children declined, and spending on the elderly grew modestly. They concluded that “during times of economic downturn or stagnation combined with government fiscal crisis…programs for children are hit much harder by the budgetary axe than programs for the elderly.”⁹ In this study we evaluate trends in social welfare spending for children and the elderly from 1980 to 2000 and the relationship of national economic trends to social welfare spending patterns.

Study Data And Methods

Methods

Following the approach of Benjamin and colleagues, we examined the major components of public social welfare spending that directly benefit children and the elderly.¹⁰ Generally, we considered children to be those younger than age eighteen, and the elderly, those older than age sixty-five. For programs that directly benefit children, we included primary and secondary education (K–12), Aid to Families with Dependent Children (AFDC) and Emergency Assistance/Temporary Assistance for Needy Families (TANF), other child welfare, nutritional support, maternal and child health programs, medical spending under Medicaid and the State Children’s Health Insurance Program (SCHIP), and those portions of Social Security payments (Old Age, Survivors, and Disability Insurance [OASDI] and Supplemental Security Insurance [SSI]) that benefit children. For programs that directly benefit the elderly, we included medical spending under Medicare and Medicaid, Older Americans Act expenditures, veterans’ and railroad workers’ pensions, and those portions of Social Security (OASDI and SSI) and food stamps that benefit the elderly.¹¹ Private social welfare expenditures, such as those from philanthropic organizations, and indirect social welfare benefits, such as those provided through the tax system, were not included. In cases where data sources were discrepant or not up to date, we contacted the relevant government agency and individuals responsible for producing annual statistical reports to obtain the most recent and accurate data. Total social welfare spending was calculated as the sum of all relevant components for both groups.

Expenditures were adjusted to 2000 U.S. dollars using the medical component of the Consumer Price Index (CPI) for Medicaid, Medicare, and SCHIP and the overall CPI for all other programs.¹² National economic trends were classified as follows: 1980–1984 recession, 1985–1989 growth, 1990–1994 recession, and 1995– 2000 expansion. Unlike expansion, inflation is present during periods of growth. To relate changes in per capita spending to national economic trends, percentage changes were calculated for each time period.¹³ Per capita social welfare spending for children and that for the elderly were analyzed separately.

Data sources

Data on primary and secondary education are from the 2001 Digest of Education Statistics and the National Center for Education Statistics.¹⁴ Data on TANF/AFDC, Social Security, SSI, and veterans’ pensions are from the U.S. Department of Health and Human Services (HHS) Social Security Administration and Department of Veterans Affairs.¹⁵ Railroad retirement and Older Americans Act data are from the U.S Railroad Retirement Board and volumes compiled for the U.S. House of Representatives Committee on Aging.¹⁶ Data on food stamps; Women, Infants, and Children (WIC); and child nutrition are from the U.S. Department of Agriculture. ¹⁷ Data on other maternal and child health programs and child welfare are from the HHS Administration for Children and Families. Medicaid, SCHIP, and Medicare data are from the Centers for Medicare and Medicaid Services (CMS). Poverty and population data are from the U.S. Bureau of the Census. CPI, medical CPI, and GDP data are from the Bureau of Labor Statistics.

Findings

Distributional spending

The distribution pattern of social welfare spending on children and the elderly changed little from 1980 to 2000 (Exhibit 2). Education constituted the bulk of public spending for children over this period; it grew from 70.3 percent of total social welfare spending on children in 1980 to 81.2 percent in 2000. Children’s income assistance provided through AFDC (renamed TANF in 1996) represented the second largest portion. The decline in the relative proportion of funds devoted to income assistance and in the real dollar value of total spending in this area occurred predominantly after the passage of the 1996 PRWORA (Exhibit 3).¹⁸ There were declines in total food-stamp spending on children after 1996 and in the real dollar value of monthly food-stamp benefits. For the elderly, Social Security and Medicare represented the majority of social welfare spending distribution from 1980 to 2000, and spending in both areas grew consistently.

EXHIBIT 2.

Distribution Of Social Welfare Spending For Children And Elders, 1980 And 2000

	Spending on children (%)			Spending on elders (%)
Program	1980	2000	Program	1980	2000
Lower education	70.3	81.2	Social Security	58.7	61.7
TANF/AFDC and Emergency Assistance	9.1	2.3	SSI	1	0.6
Social Security dependent benefits (OASDI)	1.4	1	Railroad retirement	2.5	1.3
SSI for children 0–17	0.3	1.1	Veterans’ pension	1.7	1.1
Food stamps	4.5	2.4	Food stamps	0.5	0.2
Child nutrition (school food programs)	2.6	1.6	Medicare	27	29.7
WIC	0.5	0.9	Medicaid	8.1	5.2
Medicaid	3.9	5	Older Americans Act	0.5	0.2
SCHIP	0	0.6
Maternal and child health	0.7	0.6
Child welfare	0.6	0.1

SOURCES: Authors’ analyses of data from the following sources: Data on primary and secondary education are from the 2001 Digest of Education Statistics and the National Center for Education Statistics. Data on Temporary Aid for Needy Families (TANF)/Aid to Families with Dependent Children (AFDC), Social Security, Supplemental Security Income (SSI), and veterans’ pensions were obtained from the U.S. Department of Health and Human Services (HHS), Social Security Administration and the U.S. Department of Veterans Affairs. Data on railroad retirement and the Older Americans Act were obtained from the U.S Railroad Retirement Board and volumes compiled for the U.S. House of Representatives Committee on Aging. Data on food stamps; Women, Infants, and Children (WIC); and child nutrition are from the U.S. Department of Agriculture. Data on other maternal and child health programs and child welfare were obtained from HHS, Administration for Children and Families. Data on Medicaid, the State Children’s Health Insurance Program (SCHIP), and Medicare were obtained from the Centers for Medicare and Medicaid Services. Full bibliographic information is available in an online appendix, content.healthaffairs.org/cgi/content/full/23/5/131/DC1.

NOTES: Figures are expressed as a percentage of total social welfare spending on children or elders in a given year. For example, education accounted for 70.3 percent of total social welfare spending on children in 1980. OASDI is Old Age, Survivors, and Disability Insurance.

EXHIBIT 3.

Spending On Social Welfare Programs For Children And Elders, Millions Of Dollars, Selected Years 1980–2000

Children	1980 ($)	1984 ($)	1988 ($)	1992 ($)	1996 ($)	2000 ($)
Primary/secondary education	197,385	206,106	244,778	286,692	315,334	372,865
TANF/AFDC and Emergency Assistance	25,417	23,714	24,291	26,467	22,594	10,490
Social Security (OASDI)	21,421	17,415	16,368	16,607	18,687	19,335
SSI	758	961	1,457	3,280	4,979	4,866
Food stamps	12,659	10,183	12,646	20,822	19,576	11,106
Child nutrition	7,369	6,074	6,108	6,563	7,113	7,557
WIC	1,483	2,270	2,595	3,174	4,037	3,981
Medicaid	10,970	9,907	11,187	20,248	20,099	23,000
SCHIP	0	0	0	0	0	2,781
Maternal and child health	1,817	1,753	2,445	2,574	2,480	2,654
Child welfare	1,630	270	346	335	304	292
Total social welfare for children	280,909	276,899	322,219	386,759	415,205	458,927
Elders
Social Security (OASDI)	223,300	269,617	296,903	332,712	361,486	388,096
SSI	3,694	3,277	3,492	3,604	3,636	3,775
Railroad retirement	9,638	9,973	9,637	9,403	8,893	8,303
Veterans’ pension	6,529	7,489	6,689	6,915	6,731	6,774
Food stamps	1,738	937	1,113	1,548	1,529	1,092
Medicare	102,651	129,758	138,090	148,447	165,526	186,649
Medicaid	30,696	31,906	32,778	40,630	42,327	33,000
Older Americans Act	2,023	1,838	2,220	1,519	1,315	1,284
Total social welfare for elders	380,270	454,795	490,924	544,778	591,444	628,973
Total social welfare—children and elders	661,179	731,694	813,143	931,538	1,006,649	1,087,899
GDP	5,695,740	6,430,136	7,374,289	7,722,592	8,563,592	9,872,900

SOURCES: Authors’ analysis based on data from the following sources: Data on primary and secondary education are from the 2001 Digest of Education Statistics and the National Center for Education Statistics. Data on Temporary Assistance for Needy Families (TANF)/Aid to Families with Dependent Children (AFDC), Social Security, Supplemental Security Income (SSI), and veterans’ pensions were obtained from the U.S. Department of Health and Human Services (HHS), Social Security Administration and the U.S. Department of Veterans Affairs. Data on railroad retirement and the Older Americans Act were obtained from the U.S Railroad Retirement Board and volumes compiled for the U.S. House of Representatives Committee on Aging. Data on food stamps; Women, Infants, and Children (WIC); and child nutrition are from the U.S. Department of Agriculture. Data on other maternal and child health programs and child welfare were obtained from HHS, Administration for Children and Families. Data on Medicaid, the State Children’s Health Insurance Program (SCHIP), and Medicare were obtained from the Centers for Medicare and Medicaid Services. Poverty and population data were obtained from the U.S. Bureau of the Census. Gross domestic product (GDP) data were obtained from the Bureau of Labor Statistics. An expanded version of this exhibit is available as Expanded Exhibit 3, content.healthaffairs.org/cgi/content/full/23/5/131/DC1, along with an online appendix containing full bibliographic information.

NOTES: Spending is reported in 2000 constant dollars. Numbers may not add exactly to reported total and subtotals because of rounding. OASDI is Old Age, Survivors, and Disability Insurance.

“Real” spending and as percentage of GDP

Social welfare spending in constant dollars (adjusted for inflation) for both children and the elderly grew over the study period. For children, real spending (in 2000 dollars) grew from $281 billion in 1980 to $459 billion in 2000. For the elderly, real spending grew from $380 billion in 1980 to $649 billion in 2000. However, as a percentage of GDP, social welfare spending for both groups was relatively unchanged (Exhibit 4).

EXHIBIT 4.

Social Welfare Spending On Children And Elders, As Percentage Of GDP, 1980–2000

SOURCE: Authors’ analysis of data on a variety of social welfare components. See Exhibit 3 sources and an online appendix, content.healthaffairs.org/cgi/content/full/23/5/131/DC1, for full bibliographic details.

NOTE: GDP is gross domestic product.

Per capita spending

We next examined trends in per capita social welfare spending to account for differences in population growth. Per capita social welfare spending grew from $4,464 per child in 1980 to $6,380 per child in 2000 (Exhibit 5). Per capita spending for the elderly grew from $15,404 in 1980 to $19,702 in 2000. The gap in per capita spending between the two groups grew nearly 20 percent—from about an $11,000 difference in 1980 to a more than $13,000 difference in 2000.

EXHIBIT 5.

Per Capita Social Welfare Spending On Children And Elders, 1980–2000

SOURCE: Authors’ analysis of data on a variety of social welfare components. See Exhibit 3 sources and an online appendix content.healthaffairs.org/cgi/content/full/23/5/131/DC1, for full bibliographic details.

NOTE: Changes in per capita spending for each time interval are presented in 2000 constant dollars. Changes in per capita spending were calculated as A–B. where A is per capita spending in year 2 and B is per capita spending in year 1.

Differences in per capita spending between children and the elderly grew during the recessions of the early 1980s and the early 1990s. As previously demonstrated by Benjamin and colleagues, spending on children’s programs suffered the most in 1980–1982 during the recession, when there was a 5.2 percent decrease in real total per capita total social welfare spending.¹⁹ In the ensuing period of growth, per capita public spending trends converged because of increased growth in spending on children. Real total per capita social welfare spending grew 15.8 percent for children and only 0.2 percent for elders. Despite this overall growth in per capita spending for children, most areas of social welfare spending on children declined in real dollars.

Increases in spending on children through Medicaid and food stamps made up for declines in other areas. In the recession of the early 1990s, differences between per capita public spending on children and the elderly increased as growth of spending on children failed to keep pace with that for the elderly. During this recession, real total per capita total social welfare spending on children grew 6.0 percent; on elders, 7.7 percent.

During the recent economic expansions of the late 1990s, total per capita social welfare spending rose 9.4 percent for children and 2.9 percent for the elderly. However, spending on children through AFDC and TANF dropped precipitously after passage of the 1996 PRWORA. At the same time, although medical spending on the elderly kept pace with inflation, medical spending on children fluctuated. From 1995 to 1997 it declined 22.5 percent and then increased 13.7 percent from 1998 to 2000 as Medicaid expansions and SCHIP were implemented.

Explaining the gap

We found that differences in the growth rates of medical spending accounted for the majority of the widening gap during the study period. Per capita medical spending on children remained relatively flat, while per capita medical spending on elders grew (Exhibit 6). In contrast, the dominant mode of social welfare spending on children, per capita in-kind benefits (such as education and food assistance), kept pace with per capita spending on cash benefits (for example, Social Security and pensions) for the elderly, the predominant mode of spending on this group.

EXHIBIT 6.

Trends In Per Capita Spending On Children And Elders, 1980–2000

SOURCE: Authors’ analysis of data on a variety of social welfare components. See Exhibit 3 sources and an online appendix content.healthaffairs.org/cgi/content/full/23/5/131/DC1, for full bibliographic details.

NOTE: Cash benefits include Special Security, Supplemental Security Income, railroad retirement, veterans’ pensions, and Temporary Assistance for Needy Families (TANF)/Aid to Families with Dependent Children (AFDC) and emergency assistance. In-kind benefits include education, food assistance, Older Americans Act, child welfare, and maternal and child health programs. Medical programs include Medicare, Medicaid, and the State Children–s Health Program.

Discussion

Public policy and economic trends

Changes in public policy enacted in response to national economic trends have had great influence on the allocation of public spending.²⁰ Spending on children’s programs suffered the most between 1980 and 1982 during the recession, when there was a 5.2 percent decrease in real per capita spending. To some extent, this large reduction was attributed to an unprecedented increase in the size of the population under age eighteen (which had been steadily decreasing since 1967). Substantial cuts in income-tested programs for children were authorized by the Omnibus Budget Reconciliation Act (OBRA) and the Economic Recovery Tax Act of 1981.²¹ The result was that Medicaid and AFDC spending for children declined 10.7 percent and 7.2 percent, respectively, during this recession. Because of changes in eligibility criteria, Social Security payments to surviving children decreased even more steeply, by 19.3 percent.

The Family Support Act (FSA) of 1988 took effect just as the economic growth of the mid-1980s ended, largely in response to growing public dissatisfaction with increases in welfare case loads.²² The FSA reflected bipartisan consensus in which liberals achieved a broader safety net and conservatives achieved stronger work requirements for welfare recipients. Our analysis shows that social welfare spending on poor children through AFDC peaked in 1993 and then steadily declined in large part because of implementation of the FSA. The FSA served as a precursor to the dramatic shift in social policy that occurred with passage of the 1996 PRWORA. Although this legislation decoupled Medicaid and food-stamp eligibility from eligibility for cash assistance to preserve these benefits for needy families, multiple reports at the national, state, and local levels documented enrollment declines in all three programs.²³ Our analysis confirms that declines in welfare case-loads were accompanied by declines in total public spending in TANF, Medicaid, and food stamps. Subsequent efforts to remedy inappropriate disenrollment from Medicaid, combined with the movement to Medicaid managed care, Medicaid privatization, and implementation of SCHIP, led to increased outreach efforts to enroll all eligible children in Medicaid.²⁴ As a result, spending levels returned to their 1993 levels by 1999.

Devolution to the states

As the U.S. population ages and elderly people continue to have increasing and unmet medical needs, our observation that the growth of per capita medical spending on the elderly has exceeded the growth of per capita medical spending on children is not surprising. In an attempt to preserve health care benefits for elderly people in the future, the Balanced Budget Act (BBA) of 1997 included provisions to ensure the solvency of federally guaranteed Medicare trust funds.²⁵ In contrast, current budgetary proposals continue to transfer funding responsibility and determination of eligibility requirements for social welfare programs, including Medicaid, from the federal government to the states, discounting the needs of 8.5 million uninsured U.S. children.²⁶ This trend is particularly alarming in light of the fact that children represent an increasing proportion of the population living in poverty and that poverty has been persistently and repeatedly shown to be associated with poor health outcomes.²⁷ Moreover, given that a large proportion of adult morbidities, such as cardiovascular disease and cancer, are associated with preventable childhood precursors, such as obesity and smoking, increasing and stabilizing investments in child health merit serious consideration.²⁸

Study limitations

This study has several limitations. First, some programs providing benefits to both groups, such as public housing and the low-income home energy assistance program, do not provide data classified by age. These programs were not included in our analysis. Second, our analysis did not include indirect social welfare benefits for children and elders, such as those provided through tax subsidies. The Earned Income Tax Credit (EITC), a refundable federal income tax credit for low-income working individuals and families, underwent modifications in 1993 and 1996 that led to large increases in the number of recipients and refunded amounts.²⁹ Even though the total amount refunded grew from $1.3 billion in the 1980s to more than $26 billion in 2000, this would not offset the $190 billion gap in social welfare spending between children and the elderly in 2000.³⁰ Furthermore, to avoid bias, consideration of tax policies designed to benefit children (such as refundable credits and nonrefundable deductions)warrants consideration of tax policies designed to benefit the elderly (such as tax-deferred retirement accounts and the Tax Credit for Elderly and Disabled).³¹ Estimating benefits to the elderly through the tax system would require complex statistical modeling that relied on untested assumptions about retirement account contributions, earnings, and distributions. Taking these issues into consideration, indirect social welfare benefits through the tax system were not included in our analysis. Further research in this area is needed. Despite these limitations, our findings provide evidence that unlike spending on elders, social welfare spending on children is vulnerable to downturns in the U.S. economy.

Persistent disparities in the proportion of children versus elders living in poverty and the importance of investments in preventive health argue for increased and stable social welfare spending on children’s programs. The continued transfer of responsibility for funding of social welfare programs from the federal government to the states endangers the stability of funding for all children’s social welfare programs, especially health insurance. Whereas the elderly are afforded a basic guarantee of support regardless of economic and political changes, social welfare spending on children is left vulnerable to these fluctuations. In the current period of economic stagnation, U.S. spending on children’s social welfare programs and health care is likely to be further compromised.