Future Cost Trends On Government Aid to the Poor

[Note: This is testimony by Jeffrey Kling, Associate Director for Economic Analysis (Congressional Budget Office), before the Subcommittee on Human Resources, Committee on Ways and Means, U.S. House of Representatives.]

The federal government devotes roughly one-sixth of its spending to 10 major means-tested programs and tax credits, which provide cash payments or assistance in obtaining health care, food, housing, or education to people with relatively low income or few assets. Those programs and credits consist of the following:

• Medicaid,
• The low-income subsidy for Part D of Medicare (the part of Medicare that provides prescription drug benefits),
• The refundable portion of the earned income tax credit (EITC),
• The refundable portion of the child tax credit,

• Supplemental Security Income,
• Temporary Assistance for Needy Families,
• The Supplemental Nutrition Assistance Program (SNAP, formerly called the Food Stamp program),
• Child nutrition programs,
• Housing assistance programs, and
• The Federal Pell Grant Program.

In 2012, federal spending on those programs and tax credits totaled $588 billion. (Certain larger federal benefit programs, such as Social Security and Medicare, are not considered means-tested programs because they are not limited to people with specific amounts of income or assets.)

Total federal spending on those 10 programs (adjusted to exclude the effects of inflation) rose more than tenfold—or by an average of about 6 percent a year—in the four decades since 1972 (when only half of the programs existed). As a share of the economy, federal spending on those programs grew from 1 percent to almost 4 percent of gross domestic product over that period. (For ease of presentation, this testimony uses the term “programs” to encompass both the spending programs and the tax credits.)

Medicaid accounted for more than 40 percent of the federal spending on those programs in 2012, followed in size by SNAP. A decade from now, Medicaid will account for an even larger share of spending on those programs, CBO projects. A new means-tested program—federal subsidies to help low- and moderate-income people buy health insurance through insurance exchanges, which will begin in 2014—will be the second-largest means-tested program in 2023, CBO estimates.

The projections of federal spending in this testimony are based on CBO’s current-law baseline published in February 2013. CBO updated those baseline budget projections in May 2013. Incorporating the May revisions into the numbers presented here would not change the basic findings of CBO’s analysis.

What Caused Total Spending on Means-Tested Programs and Tax Credits to Rise Over the Past 40 Years?

Two broad factors were responsible for the growth of spending on means-tested programs and tax credits between 1972 and 2011: increases in the number of people participating in those programs and increases in spending per participant. (This discussion focuses on the 40-year period ending in 2011 because that was the most recent year for which data on the number of participants were available for those programs.) Both of those increases were themselves the result of multiple factors. For example, the rise in participation stemmed from three important causes:

• Population growth (the U.S. population increased by almost 50 percent during that period),
• Changes in economic conditions (particularly the recession that occurred from 2007 to 2009 and the weak recovery that followed it), and
• Actions by lawmakers to create new means-tested programs and tax credits and to expand eligibility for some existing ones.

Increases in spending per participant resulted mainly from two factors:

• Growth in the cost of providing assistance (such as rising costs for medical care), and
• Actions by lawmakers to provide more generous benefits (such as increases in SNAP benefits).

What Caused the Growth of Specific Categories of Means-Tested Programs and Tax Credits?

Roughly half of the total growth in spending between 1972 and 2011 on the means-tested programs and tax credits examined in this testimony came from increases in spending for health care programs—Medicaid and, to a far lesser extent, subsidies to help low-income people pay for prescription drugs under Part D of Medicare. Spending for such programs grew about 15-fold over the 1972–2011 period: from $20 billion to $305 billion. (Those and other dollar amounts for program spending in this testimony are presented in 2012 dollars to remove the effects of inflation.) The main reason for that growth—which averaged about 7 percent a year above the rate of inflation—was increased spending per participant. Had the amount of spending per participant in Medicaid remained unchanged between 1972 and 2011, total spending on the health care programs examined here would have been about $88 billion in 2011, or less than one-third of the actual amount.

Growth in each of two other broad categories of means-tested programs and tax credits—programs that provide cash assistance and programs that help people obtain food, housing, or education—was about equally responsible for the other half of the total increase in spending over the 1972–2011 period on the programs included in this testimony:

• Spending on cash assistance programs and tax credits (the largest of which is the refundable portion of the EITC) rose from $18 billion in 1972 to $151 billion in 2011—that is, by nearly 6 percent a year above the rate of inflation.
• Spending on programs that help people afford food, housing, or education (the largest of which is SNAP) rose from $17 billion to $172 billion—that is, by about 6 percent per year above the rate of inflation.

Unlike growth in spending for health care programs (primarily Medicaid), which stemmed mainly from greater spending per participant, growth in spending for those other programs resulted primarily from increases in the number of participants.

How Much Will Means-Tested Programs and Tax Credits Cost Over the Next Decade?

If current laws that govern the means-tested programs and tax credits examined in this testimony do not change, total spending on those programs will grow faster than inflation over the next decade, CBO projects. However, changes in spending will vary among programs.

Spending on means-tested health care programs is projected to more than double, from $272 billion in 2012 to $624 billion in 2023 (adjusted for inflation), an average annual increase of 8 percent above the rate of inflation. That rise reflects expected growth in the cost of providing medical care; it also reflects expanded eligibility for assistance and new types of assistance to be provided under the Affordable Care Act (ACA). The ACA will not only make more people eligible for Medicaid but also allow many low- and moderate-income people who do not qualify for Medicaid to purchase federally subsidized health care coverage. Subsidizing that non-Medicaid coverage through cost-sharing subsidies and the refundable portion of a new premium assistance tax credit is projected to cost $109 billion (in 2012 dollars) in 2023.

In contrast to spending on health care programs, total spending on the cash assistance programs and tax credits examined here is projected to fall over the next decade—from $148 billion in 2012 to $137 billion in 2023 (adjusted for inflation). That expected decline mainly stems from changes to the earned income and child tax credits that are scheduled to occur under current law.

CBO also estimates that spending on the nutrition and education programs discussed here will decline in the next 10 years, partly because spending on SNAP is projected to drop substantially as the economy continues to recover. (Unlike the other programs included here, the housing assistance programs rely on annual appropriations for all of their funding. In this testimony, CBO has not projected the size of those appropriations, which will depend on future actions by lawmakers.)