While my prior post organized public expenditures by category of public service, there is another way to look at it – by basis of eligibility. As the data in the prior post showed, only about 20% of public spending went for general services for everyone, with most of the rest either payments for the past (interest, pensions) or services and benefits with eligibility limitations. The most common limitations are based, at least in theory if not always in practice, on age, means (the amount of money one has or earns), needs (disability), or some combination of these. Spending by basis of eligibility is analyzed, using a variety of data from the 2001 Statistical Abstract, in the attached spreadsheet. Unlike the 1995 data previously prevented, this table is a hodgepodge and must therefore be thought of as a rough estimate. The tabulation finds that programs with eligibility limited by age, means and need accounted for about half of all government expenditures in 1998, with age-restricted benefits accounting for the vast majority.
It is not unreasonable to characterize the government as an entity that collects resources from working adults and redistributes it to the elderly and children. A family may be similarly characterized, so the government may be thought of as a compliment to, or a substitute for, the family. Where family obligations end and community obligations begin is the subject of legitimate differences of opinion. Often the community steps in to assist needy children and seniors whose related adults are not providing for their needs. This, to some, is unfair to those adults who do provide for their family members, and are then taxed to pay for others. I call this the “family of last resort dilemma.”
Note that the spending in the spreadsheet is “direct spending,” tabulated for the level of government that actually pays out. A significant share of state and local spending in age, means and needs categories is (as we will see later) federally financed.
Although in theory anyone can enroll in public universities and colleges, in practice their benefits accrue primarily to the young. Only the young are eligible to attend public elementary and secondary school, and only the old are eligible to receive retirement benefits under social security, and Medicare. Taken together, major categories of services with eligibility limited based solely on age accounted for 32 percent of all government spending in 1998. Smaller age-based programs that are lumped in with other categories, such as community centers for the elderly that do not restrict eligibility based on means, and the cost of half-fare discounts for the elderly on mass transit, would push that share higher, perhaps to a third.
There were 281 million Americans in the year 2000, with 35 million age 65 or over and 99 million ages 24 or under (2001 Statistical Abstract, Table 11). In that year, about 39 million received age-based social security benefits (Table 527); this included just about everyone receiving retirement earnings. In addition, 39 million were enrolled in Medicare (Table 136). Just over 47 million attended public schools, or 89 percent of those enrolled in school; and 11.8 million attended public colleges and universities, about 78 percent of those enrolled in colleges and universities (Table 207). Therefore, about 35 percent of all Americans – but a far higher share of those in the eligible age groups – benefited from these public services. Within their categories, they are nearly universal.
An additional set of services are available only to those who are old or young, but are in addition “means tested” to limit availability to those with lower incomes. Those who have more income or wealth are expected to pay for these services themselves, rather than receiving public funding. For the elderly, the most important means-tested benefit is Medicaid, the health care program for the poor. Medicaid expenditures on elderly recipients accounted for 1.4 percent of all government spending. It would be more today, and will be more tomorrow. Since home health care and nursing home care is so expensive, and is only covered by Medicare for a limited period of time, most elderly people requiring those services eventually qualify for Medicaid even if they had not previously been poor. The poor elderly are also eligible to receive food aid, such as subsidized meals on wheels. The most important services for poor children are foster care, school breakfasts and lunches, pre-school funded through Head Start, subsidized child care, and student loans and grants for college. Together, programs with eligibility based on both age and means accounted for 2.7 percent of expenditures.
In 1998 there were 12.8 million children (Table 680), and 3.4 million elderly people, living in poverty. During that year, just under four million elderly people were on Medicaid, or about one of every nine people over age 65, and more than the number of elderly persons living in poverty (Table 139). Elderly people may also qualify for Medicaid by being “medically needy,” that is having medical expenditures that exceed a certain percent of their income. Although they accounted for less than 10 percent of all Medicaid recipients, the elderly accounted for 29 percent of all Medicaid expenditures, since the medical services they require are typically extensive. In addition, elderly people who never worked, but instead “retired” from welfare and receive Supplementary Security Income (SSI) are not eligible for Medicare, but receive Medicaid instead. This is significant because state and local governments pay for a substantial share of Medicaid expenditures, while the federal government fully funds Medicare. As the “welfare” generation hits age 65, the number of elderly people on SSI and Medicaid is rising, even as the number of people on welfare falls. The most important fact about Medicaid for the elderly, however, is that the number of elderly people receiving Medicaid exceeded the number of elderly people who were poor.
There were 15.3 million children in the school lunch program and 7.4 million in the Women Infant Children program (WIC); therefore about one child in five received food aid over and above food stamps (Table 524). Again, this is somewhat more than the number of children living in poverty according to survey estimates. There were 306,000 children in foster care, and 822,000 in head start, a small share of the young in each case. Among those in college, 3.7 million received Pell Grants, 5.0 million received Stafford Loans, and about a million were in federal work study programs. According to another table (272), 19 percent of all college students received federal Pell Grants, 19 percent received state grants, and 28 percent received federally guaranteed student loans. Additional need-based grants and scholarships by foundations and colleges themselves generally use the same means testing formula as the federal government. Studies have found that only one of every seven children whose parent’s means qualifies them for subsidized pre-school child care actually receives it.
Another set of programs provides assistance to households based on their means alone, providing either cash assistance or goods and services to households headed by poor and unemployed adults. Cash assistance programs, including the federal Earned Income Tax Credit, unemployment compensation, and welfare payments, accounted for 2.3 percent of government expenditures. In-kind assistance programs, primarily for food, housing, and health care, cost much, much more, accounting for 7.1 percent of total government expenditures. Health care accounted for most of this: Medicaid for those with low incomes, rather than the aged or disabled, accounted for 1.6 percent of all expenditures, while public health and hospitals, which are primarily used by the poor, accounted for 3.9 percent. Other major programs include food stamps (0.8 percent), and “Section 8” rent subsidies for the poor (0.6 percent).
While these categories of expenditures are based on means, not on age, and provide assistance to entire households, many of the beneficiaries are children. There were 32 million U.S. residents below the poverty level in 1999, according to the Census Bureau, and 44 million below 125 percent of the poverty level (Table 679). Excluding the elderly, there were 28 million people in poverty. In 1998, there were 30 million non-elderly, non-disabled adults and children on Medicaid (Table 139), about one of every eight U.S. residents under age 65 and over age 17, and about the same number as the population in poverty. The number would be higher today, as private health insurance shrinks.
According to another table, however, (Table 137), only 40 percent of those living in poverty received Medicaid, and those in poverty accounted for less than half of all Medicaid recipients. While accounting for three-quarters of all Medicaid recipients, the non-elderly non-disabled accounted for just 29 percent of all Medicaid expenditures, since their health care costs tend to be lower. Also in 1998, 21 million people received food stamps, 3.0 million received housing assistance, and 1.3 million resided in public housing. Therefore, while in-kind health care and food aid programs cover a substantial share of the poor, in-kind housing programs only help a small, fortunate minority (Table 524).
As for cash assistance, 8.8 million received Temporary Assistance for Needy Families, the federal program most people think of as “welfare.” That number is far lower today. In 1997, an additional 700,000 people not eligible for federal welfare received benefits under state general assistance programs. On the other hand, 58 million working poor people received at least some money from the Earned Income Tax Credit in 1998 (all Table 524). In 1999, an average of 2.2 million workers received state unemployment insurance payments (Table 537); these workers merely have to be unemployed, not poor. But there was an average of 5.9 million unemployed workers actively looking for work that year (Table 569).
Finally, another group of public programs limits eligibility based on both means and need, with need measured by poor physical or mental health or social problems. With the exception of worker compensation those who have equal needs, but who are not poor, are expected to pay for similar services themselves. In theory any health care program, including Medicare and Medicaid for the elderly and poor, can be so characterized, since services are theoretically only provided to those who are sick. The programs in this group, however, provide income and health care for working age adults who are permanently disabled, and who require a public determination of disability.
The federal Supplemental Security Income program provides income to the disabled and their dependents; excluding payments to the elderly, it accounted for 1.4 percent of total expenditures. Worker compensation payments accounted for another 0.6 percent. Medical care for the disabled adults, provided under Medicaid, accounted for an additional 2.1 percent of government expenditures, a greater share than that of cash income under SSI and worker compensation. Finally, social services accounted for the poor and troubled accounted for just two tenths of one percent of overall expenditures. Altogether, services with eligibility limited by need and means accounted for 4.3 percent of total government expenditures.
About 17 million U.S. residents ages 16 to 64 self-identified themselves as having a work disability in the year 2000, according to a Census Bureau survey (Table 591). Of these, only 4.7 million were employed, with another half million actively looking for work (an unemployment rate of 9.8 percent). The rest – 12.3 million – were out of the labor force entirely. In 1998, 4.7 million disabled workers and 2.0 million of their dependents, plus 2.0 million widowed mothers and children of deceased workers, received Supplemental Security Income (SSI) (Tables 527 and 528). According to another table (538), only 32 percent of disabled workers received SSI. Slightly more than 1.8 million disabled workers received worker compensation. In 1998, 6.6 million people received Medicaid on the basis of being disabled; though accounting for only 16 percent of all Medicaid recipients, the disabled accounted for 42 percent of all Medicaid spending, since they have extensive health problems. There were 261,900 beds in facilities for the mentally ill and retarded in 1998 (Table 177).
Not everyone is eligible for the public services and benefits that have been discussed here. Among those theoretically eligible, not everyone receives them, as comparisons between recipients and those in various demographic categories makes clear. The rules and practice for determining eligibility are frequently opaque and poorly understood. But behind them are all kinds of moral judgments, economic principles, and political deals and maneuvers. I will discuss the ethics of eligibility in an upcoming series of posts.