Executive summary
The American social safety net has historically revolved around helping low-income families with children, poor elderly adults, and disabled individuals. However, one of the largest components of the American safety net, the Social Security program, is often overlooked as a mechanism to help children and young adults. The Social Security program already helps millions of children via survivor benefits, retiree child benefits, and disability benefits. Yet, there remain large gaps in the child safety net, especially as the needs of families continue to evolve. This issue brief is part of a larger body of work entitled Promoting Economic Security for Older Adults that will strengthen federal policies and restore solvency to key social safety net programs for the elderly. In this brief, we will focus on Social Security benefits for children.
We propose reforming the Social Security program to provide additional support to low-income grandparent caregivers and students; we propose offsetting these costs by eliminating spending on child benefits for retirees. The first reform is a benefit for grandparents raising grandchildren.1 This new caregiver benefit for low-income grandparents with dependent grandchildren will help to address the consequences of increased rates of grandparent caregiving by providing financial resources to grandparent families. The second reform restores a benefit for students who are children of deceased or disabled workers. The benefit, originally eliminated in 1982, will provide essential support to dependents who wish to attend vocational training or college after secondary school. Finally, we propose the elimination of child benefits for retirees. Currently, this benefit primarily supports older men with limited financial need.
Low-income caregiving grandparents should receive additional support
An increasing number of children live in multigenerational and skipped generation households. The American Community Survey (ACS) reports 4.8 million multigenerational households in 2023, while the Current Population Survey (CPS) reports 3.7 million multigenerational households (Table 1). This amounts to about 3.4% (ACS) of all American households or 12.5% (ACS) of all households with children.2 Of these multigenerational households, about half of them report relying on grandparents for at least 50% of a dependent’s financial support. While many of these households are missing the grandchild’s parents (a.k.a “skipped generation”), a majority have a parent present but rely on the grandparent(s) for financial support.
Grandparents are most often placed into a caregiving role due to divorce, substance abuse, incarceration, or mental health issues. The child welfare system has moved from placing children who cannot be safely cared for by their parents in foster care settings to kin-care settings. This policy change has proven to be beneficial for the children and can be further enhanced through improved assistance to the grandparents raising grandchildren. Additionally, grandparent caregiving is rarely planned for. Grandparents are often asked to step in without warning and are suddenly faced with the unexpected costs of one or several children. Thus, while kin care is better for the child, it should still be recognized that these households are typically formed in response to adversities that make the children and grandparents highly susceptible to further instability.
Since incarceration and substance abuse disproportionately affect low-income and minority households, the burden of grandparent caregiving is also largely in low-income and minority communities. Subsequently, grandparent caregivers also experience worse health and financial outcomes compared to the overall elderly population. They have significantly less wealth, receive less retirement support, and report higher rates of poverty after retirement compared to the general elderly population. They are also younger but report higher rates of chronic conditions and lower well-being. Grandparent caregivers face considerable challenges, but they continue to play a critical role in the lives of their grandchildren while providing relief to the formal foster care system.
Current support for caregiving grandparents is insufficient, although some support is available through Temporary Assistance for Needy Families (TANF). Financial support for caregiving grandparents primarily comes from the tax system. The tax code allows grandparents to claim their grandchildren as dependents if the child lives in the grandparent’s home for 50% of the year and the grandparent provides over half of their financial support. However, these benefits do not reach individuals who do not file taxes. Additionally, child assistance through the tax code is insufficient. The maximum amount of the child tax credit that is refundable was $1,700 per child in 2025, while the incremental cost to families of adding a child to a household is $5,500 (according to federal poverty guidelines). Other social welfare programs, including programs like Social Security and TANF, leave out caregiving grandparents almost entirely. Social Security will only pay child benefits to the grandparent if the biological parents are deceased or disabled, or if the grandparent legally adopts the child. However, most caregiving from grandparents occurs outside of the foster care or adoption system, often because the grandparents only expect temporary responsibility for their grandchild.
Thus, we propose implementing a new caregiver benefit for low-income grandparent caregivers. Grandparents would be eligible to receive caregiver benefits if they provide at least 50% of the child’s financial support, and they expect to take care of the child for at least 3 months. This caregiving benefit would provide grandparents earning less than 133% of the 3-person household federal poverty line (FPL) an additional $5,500 (in 2025 dollars) from Social Security per grandchild. The maximum per-grandchild benefit amount in a given year is determined by the amount that the FPL (at 100%) increases for each additional person in a household. This benefit would then phase out—reduced by 40 cents for every dollar of additional grandparent income beyond 133% of the 3-person household FPL. With this formula, the benefit would cap out for grandparents that earn 180% of the FPL for 3-person households (approximately $48,000 in 2025).
Households eligible for the benefit would apply at the Social Security office. Grandparent income includes adjusted gross income (which includes taxed Social Security income) plus untaxed Social Security income and Social Security Income (SSI) benefits for all grandparent caregivers in the household. An individual grandparent caregiver’s benefit amount would be recalculated annually to account for changes in income. Grandparents who are not currently receiving Social Security benefits or are below age 62 are also eligible for the benefit, as long as they have sufficient work history to be eligible for disability benefits. These benefits would not impact Medicaid eligibility.
Using the ACS3, we estimate that providing a caregiver benefit for low-income parents would cost $64 billion from 2027 to 2035.4 We also assume 90% take-up because grandparents would need to apply for the benefit. In the year 2027, the first year of the benefit, the caregiver benefit would support over 736,000 households, consisting of almost 1 million grandparents and 1.3 million grandchildren. On average, grandparents who receive the caregiving benefit would see an additional $8,752 per year (in 2027 dollars), which is equivalent to a 21% increase in their income. It would provide the greatest average subsidy to beneficiaries in the first quintile, providing them with an average $10,110 additional dollars per household in annual income. Grandparents in the first quintile tend to have a larger average benefit because they often have more grandchildren under their care, along with a reduced phase-out of their benefits. The program would amount to an additional $6.4 billion across all grandparent caregivers in 2027. The grandparent caregiver benefit would also largely benefit households with women and Black grandparents, providing, on average, $8,795 to 656,000 households and $9,016 to 188,000 households, respectively.
We propose a caregiver benefit administered through the Social Security system for several reasons. The child welfare system is state-based and resource-constrained. Providing benefits through a state system would lead to varying benefit amounts, with many states unable to provide any assistance. Social Security has efficient administrative procedures that ensure a benefit can be provided uniformly. While our proposed benefit would be limited to low-income individuals, it aligns with the existing progressivity of the Social Security program, while avoiding the stigma of a separate means-tested program. Additionally, other “universal” programs like Medicare require increased contributions from higher-income beneficiaries and subsidize benefits for lower-income beneficiaries. This income limit is also essential to maintaining fiscal responsibility—an income-blind benefit would cost $230 billion through 2035 (3.6 times our proposal).5 Ultimately, administering a low-income caregiver benefit through the Social Security program offers ease of accessibility, uniformity, and funding availability while also avoiding escalating costs.
Students should remain eligible for child benefits after secondary school
In addition to failing grandparents, the current structure of Social Security child benefits also fails to appropriately support students. When Social Security was originally established, the children of beneficiaries could only receive Social Security payments until they were 18. However, as more Americans began to pursue higher education, legislators recognized that many full-time students were dependent on their parents past the age of 18. Thus, in 1965, the definition of dependent was broadened to include full-time students under the age of 22. These benefits were provided to offset the loss of income if a worker retired, passed away, or became disabled. It was paid out per dependent child. In its peak pay-out year, 1981, $8.5 billion (in 2025 USD) was provided to more than 760,000 full-time students. However, in 1981, policymakers reformed the program to only provide benefits for students in secondary education or below, with the last benefits paid out in 1999.
Policymakers should restore this program, allowing individuals pursuing higher education, whether at a university, college, or trade school, to remain eligible for child benefits. Fundamentally, the justification for these student benefits in 1965 is still true today. Students are still largely dependent on their parent’s income. Should a student’s parent pass away or become disabled, that student would be experiencing a loss of income that, otherwise, they likely would have received. While sometimes called student benefits, these payments should be recognized as child benefits. These benefits are designed to support dependents experiencing an unexpected loss of income. The CBO, in 1978, found that student benefits tended to support lower-income families, given the economic conditions of the eligible population. The median income of student beneficiaries was 33% lower than the median income of all college students. Ultimately, this benefit made higher education, in all forms, more accessible.
Thus, we propose re-implementing the student benefit with a few adjustments—specifically, including those attending trade school and those who no longer live in the home. Students 25 years old or younger would remain eligible for child benefits while they are enrolled at least half-time in university, community college, or vocational school. This benefit would cover the biological children of deceased or disabled workers, the adopted children of retirees, or the grandchildren of low-income grandparent caregivers. To confirm enrollment, Social Security should use the same electronic enrollment verification systems as other financial aid programs like FAFSA and the Pell Grant. All students, regardless of marital status and whether they live in the home, would be eligible for this benefit.
Providing child benefits to all eligible students would cost $111.4 billion from 2027 to 2035.6 In the year 2045, after the benefit is fully implemented, we estimate that restoring student benefits would provide 492,000 individuals with additional funds for college or vocational training, amounting to a total of $4.5 billion in benefits. White female students would be most likely to claim these benefits, as this is the population most likely to pursue higher education, but 229,000 non-white students and 226,000 male students would also receive support. On average, students would receive $9,092 in additional funds annually. We also anticipate that providing a student benefit would increase enrollment and reduce attrition. Research after the 1981 reform found that the elimination of benefits for students reduced college attendance among the previously eligible population by more than 43%. Each $1,000 spent by the program increased the probability of attendance by nearly 6%.
Beyond expanding access to education, it is important that these student benefits be added specifically to the Social Security program due to the constraints on the discretionary appropriation process. Providing increased educational support via Social Security will also lessen the pressure on other forms of funding support for students pursuing higher education—most of which are determined through the appropriations process.
Eliminate child (and associated parent) benefits for retirees
While the needs of grandparents and students are neglected by Social Security, higher-income retirees with dependent minors receive billions of dollars in benefits annually. Currently, the children of retirees are eligible for 50% of a retiree’s benefits. While child benefits are a vital support for the children of disabled or deceased individuals, it is often a less meaningful benefit for the children of retirees who tend to be more financially secure.
In 2024, 323,000 children received Social Security benefits due to their parent’s retirement, and they received an average of $862 per month, amounting to $3.3 billion per year. Of the retirees receiving child benefits, 89% were men.7 Women are unlikely to have a child past the age of 45, while men do not face the same biological constraints as women. Because men can have children far later in life, they are then more likely to retire with dependent children under the age of 18. The average primary insurance amount (PIA) for retirees with dependents is $1,707, but $1,822 for all male retirees, meaning that retirees with children have a similar income to the typical American retiree.8 Having a child while older rather than younger is a personal choice, but there is no good argument for subsidizing that decision via higher Social Security payments, especially because these payments largely benefit higher-income households.
This benefit may also have downstream consequences, encouraging workers to retire early when they would not otherwise. Current policies allow individuals to receive benefits as early as age 62, but with reduced benefits. However, if a retiree has a child under age 18, that child will receive 50% of the amount the parent would have received at age 67. In addition to child benefits, retirees can also claim spousal benefits if the spouse is caring for a qualifying child. A retiree’s spousal benefits can amount to up to 50% of a retiree’s full benefit amount. Younger women are far more likely to receive spousal benefits compared to younger men. While only 91 men under the age of 62 received spousal benefits in 2024, more than 22,000 women younger than age 62 received benefits. Women entitled to spousal benefits due to children received an average monthly benefit of $811 in 2024. Considering payments to both a retiree’s children and spouse, for some families, these benefits make up for or even exceed the lost income from early retirement. Put simply, there are more equitable and more efficient ways to allocate this money to support families with greater need.
Thus, we propose eliminating child benefits for the biological children of retirees and any associated spousal benefits. Importantly, benefits would remain in place for the children of deceased or disabled parents and the adopted or disabled children of retirees. The provision would begin prospectively on January 1, 2027, meaning after this date, new retirees are no longer eligible for child and spousal benefits. Eliminating child benefits and associated spousal benefits for retired workers would lead to savings of $35.8 billion from 2027 to 2035.9
Conclusion
This brief proposed three major reforms to child benefits within the Social Security program—establishing a benefit for low-income caregiving grandparents, restoring student benefits for those pursuing higher education, and eliminating child and spousal benefits for retirees. Similar benefits have already been proposed by Representative John Larson in his 2023 Social Security proposal. The Larson proposal indicates that Congress is willing and able to address these gaps in the social safety net. By using child retiree benefits, which primarily benefit higher-income men, as a partial pay-for, Congress can prioritize families with the greatest need.
These reforms realign child benefits with their intended purpose—to support children in families with a significant loss of income. Current child benefits for retirees and their spouses function as a subsidy for older fathers. By instead spending on grandparents and students, Congress ensures that children facing lost parental income are still afforded the same opportunities as any other child. The circumstances that lead to grandparent caregiving, or a loss of parental income, are not imposed evenly. Families that are already vulnerable to financial insecurity and poor health are also the most likely to rely on these benefits. Ultimately, these reforms will ensure that children and families with the greatest need receive adequate financial protection.
-
Acknowledgements and disclosures
The authors thank Tara Watson for careful review and comments on earlier drafts. They thank Samuel Peterson for excellent fact-checking and Rasa Siniakovas for editorial assistance.
The authors thank Steve Goss, Karen Glenn, and the team in the Office of the Chief Actuary at the Social Security Administration. They also thank Karen Smith and the team at the Urban Institute.
-
Footnotes
- We refer to grandparents throughout this brief as they are most likely to provide kin-care when a parent is unable to care for the child, however, caregivers in the same generation as a potential grandparent are also eligible (such as a great aunt or great uncle).
- Author’s analysis of the ACS.
- Results from the author’s analysis of the ACS. We primarily report results from our model, which uses the ACS, because the survey allows respondents to report whether they live with a grandchild, whereas with the CPS, we must infer grandparents using interrelationship variables. As a result, we can more accurately identify grandparent caregivers in the ACS than in the CPS.
- We model the benefit using the 2023 ACS, then project costs and beneficiaries in future years by scaling the number of beneficiaries (children) by CBO Demographic Projections (https://www.cbo.gov/publication/60875) and benefit amount by CPI, assuming a 3% inflation rate for years beyond 2025.
- Author’s analysis of the ACS.
- Analysis from the Urban Institute’s Dynamic Simulation of Income Model (DYNASIM).
- Internal communication with the Office of the Chief Actuary.
- Internal communication with the Office of the Chief Actuary.
- Analysis from the Urban Institute’s Dynamic Simulation of Income Model (DYNASIM) adjusted to better account for disabled children.
The Brookings Institution is committed to quality, independence, and impact.
We are supported by a diverse array of funders. In line with our values and policies, each Brookings publication represents the sole views of its author(s).
Commentary
Improving Social Security for children and young adults in need
September 5, 2025