Children and Wealth Contexts in the United States: Differences by Household Type

Abstract

Objective:

To examine whether the wealth context of households with children, marked by high rates of inequality and low levels of wealth for those at the bottom, also applies to elderly households and households without children.

Background:

Children experience higher income poverty than elderly or working-age adults, but wealth and wealth deprivation comparisons across these groups have not been done. Exploring these differences may reveal another economic dimension on which households with children are uniquely vulnerable and inform policies aimed at financial stability.

Methods:

Data are drawn from the 1989 to 2022 waves of the Survey of Consumer Finances (N = 58,148 households), a nationally representative triannual survey of household wealth. The study tracks trends in wealth inequality, wealth holdings, and net worth poverty across three household types: non-elderly households with children, non-elderly households without children, and elderly households.

Results:

Households with children exhibit higher wealth inequality, lower wealth levels, and greater net worth poverty rates than the other two household types. Disparities between elderly and child households are particularly large, with child households having pennies on the dollar for every dollar of elderly wealth. These disparities increased over time, except in the early 2020s, when gaps narrowed.

Conclusion:

Like income, wealth is another economic context in which child households compare unfavorably to households without children and elderly households. However, government spending during the pandemic coincided with increases in child household wealth and decreases in net worth poverty, suggesting that child wealth contexts are not fixed.

In the United States, wealth—a household’s assets minus its debts—is a critical economic resource for children. Wealth predicts child well-being, including educational attainment, and these effects cannot fully be explained by income (Gibson-Davis and Hill 2021; Miller et al. 2021). As a cumulative resource that is passed on through generations, wealth has multigenerational effects, and the intergenerational transmission of wealth in the United States is high (Pfeffer and Killewald 2015; Pfeffer et al. 2016). Children who grow up in households with low wealth levels are unlikely to have sufficient wealth as adults (Pfeffer and Killewald 2018, 2019). Thus, the wealth context of children likely informs societal flourishing, both now and in the future (Gibson-Davis and Percheski 2018).

Given the importance of wealth for child well-being, recent trends in wealth and wealth inequality for US households with at least one resident younger than 18 (hereafter, child households) are alarming. Wealth inequality among child households is higher than income inequality and has been rising more rapidly (Gibson-Davis and Hill 2021). Additionally, a substantial fraction of households have little to no wealth, leading to high rates of net worth poverty. Net worth poverty refers to households whose wealth levels are less than one-fourth of the federal poverty line (in 2024, $7953 for a family of four). In 2019, the net worth poverty rate for child households was 35%, three times higher than the income poverty rate (Gibson-Davis et al. 2021). Most children who are net worth poor are not income poor, indicating that net worth poverty is a distinct marker of economic disadvantage. Net worth poverty is negatively correlated with healthy child development, an association that is independent of the negative effects of income poverty (Gibson-Davis et al. 2024, 2022).

Wealth may be an important, but heretofore unrecognized, economic dimension on which child households are worse off than other household types. Relative to elderly and working-age individuals, children in families have higher income poverty rates (Benson and Bishaw 2024; U.S. Census Bureau 2023a). This economic vulnerability of child households matters not only for the life chances of the children involved, but also for the household as a whole insofar as the government explicitly recognizes household structure in the allocation of social welfare benefits. The presence of a child, along with whether the head of the household is working age or elderly, is used to determine the scope and reach of the safety net (Moffitt 2015; Parolin, Desmond, et al. 2023). Indeed, among non-elderly households, the presence of a child in the household has become an increasingly important marker of social welfare generosity as the policy landscape has shifted to provide tax credits that are substantially more generous to households with children than those without (Parolin, Desmond, et al. 2023). As recognized by the social safety net, these three household types represent different levels of economic vulnerability, with wealth potentially representing another dimension along which child households are disadvantaged.

A particularly stark example of policy addressing the economic vulnerability of child households is US government expenditures directed toward families with children during the COVID-19 pandemic. In response to the pandemic’s onset, the government passed a series of bills—most notably, the Coronavirus Aid, Relief, and Economic Security (CARES) Act and the American Rescue Plan—totaling $5.6 trillion in tax cuts and spending provisions (Tax Policy Center n.d.). Through program outlays and tax reductions, the federal government had child-based expenditures of $834 billion in 2021, a 40% increase over 2020 amounts (Lou et al. 2022). This child-centric approach to pandemic relief successfully lowered the income poverty rate for children (Creamer et al. 2022), raising the question of whether this aid had any effect on net worth poverty rates. Indeed, the types of expenditures favored by the government—lump sum stimulus payments and tax credits—lead to wealth building among child households (Jones and Michelmore 2018; Michelmore and Lopoo 2021), suggesting that such an effect may be likely.

To understand whether wealth contexts represent another economic dimension along which child households are disadvantaged and to investigate any changes in wealth that coincided with pandemic-era spending, this study compares wealth levels and inequality across three mutually exclusive groups of households: child households (with a resident child younger than 18 and head/spouse younger than 65), non-child households (with no resident child younger than 18 and head/spouse younger than 65), and elderly households (with the head/spouse older than 65). Data come from the 1989–2022 waves of the Survey of Consumer Finances, the premier source of US wealth data. We use 35 years of data to understand any observed changes relative to longer trend changes. We compare inequality and wealth levels across the household types by examining measures of wealth inequality and differences in wealth at various points across the distribution. We also compare trends in net worth and income poverty. Analyses are stratified by year, with particular attention to wealth changes between 2019 and 2022, the period coinciding with pandemic-era spending.

This study makes three descriptive contributions. First, it provides the first evidence on whether wealth, like income, is an area where child households are disadvantaged compared to childless and elderly households. As wealth is distinct from income (correlated at 0.50; L. Keister 2018), this broadens our understanding of child household vulnerability and may inform financial security policies. Second, by comparing income and net worth poverty rates, the study expands the concept of economic deprivation to include wealth and shows how ignoring net worth can underestimate deprivation across household types. Third, using data through the early 2020s, it is the first to examine whether increased COVID-19-era government spending correlates with changes in wealth contexts. Though not causal, these results offer a critical first step in assessing shifts in children’s wealth. In sum, this comparative analysis highlights another economic margin of disadvantage for child households, potentially informing child development policy and future causal research.

1 |. Background

1.1 |. Wealth and Net Worth Poverty Among Child Households

The distribution of wealth among child households indicates that a small fraction of households has very high levels of wealth, but the majority has very little. In 2019, 10% of all child households controlled 82% of all wealth, whereas the bottom 50% had less than 0.05% (Gibson-Davis and Hill 2021). Wealth inequality among child households has been rising and has increased faster than income inequality (Gibson-Davis and Percheski 2018). The wealth context of child households is also marked by high rates of net worth poverty, with roughly one in three child households experiencing net worth poverty (Gibson-Davis et al. 2021).

Inadequate levels of wealth matter because children need sufficient levels of wealth to flourish. Theoretically, wealth is related to child well-being through several mechanisms. First, it enables parental investment—facilitating spending on childcare, schooling, home learning, and cultural activities (Gibson-Davis and Hill 2021; Grinstein-Weiss et al. 2014). It also allows families to smooth consumption during income disruptions. Parents with low wealth, as seen in net worth poverty, may lack the means to invest in their children. Second, wealth influences psychosocial stress (Conger et al. 2010; Miller et al. 2021). Low wealth can heighten parents’ feelings of subordination and lack of control (Cutler et al. 2008) and reduce their ability to manage financial risks (Barr 2012; Conger and Elder 1994). Even outside crises, low wealth may impair parent–child interactions by increasing anxiety, hopelessness, and frustration (Boen et al. 2020; Drentea and Reynolds 2015). Third, wealth reflects and shapes class status (Conley 1999), influencing how children are treated and how they see themselves, affecting behavior and academic outcomes (Destin et al. 2012; Mistry et al. 2015). Net worth–poor families may view themselves as lower class, reducing self-efficacy and social belonging (Sherraden 1991). Fourth, wealth shapes expectations about life outcomes. These expectations affect parental expectations and children’s aspirations (Destin and Oyserman 2009; Diemer et al. 2020). Low wealth may weaken parents’ future orientation and optimism about mobility (Yeung and Conley 2008), leading to underinvestment or suboptimal decisions for their children.

Consistent with these theoretical mechanisms, wealth and wealth deprivation predict educational attainment, academic achievement, and socioemotional functioning (Diemer et al. 2020; Shanks 2007; Zang et al. 2024). Wealth has particularly strong associations with educational attainment (Conley 2001; Doren and Grodsky 2016; Jez 2014; Karagiannaki 2017): children of the wealthiest parents are more than 40 percentage points more likely to graduate from college than children from the least wealthy families (Pfeffer 2018). Parental wealth is positively correlated with their children’s standardized test scores and academic achievement (Friedline et al. 2015; Miller et al. 2021; Moulton et al. 2021; Yeung and Conley 2008). Children with wealthier parents also have higher sociability and fewer behavioral problems (Berger and Houle 2019; Diemer et al. 2020; Ream and Gottfried 2019). In terms of net worth poverty, children who are net worth poor, relative to those who are not, have lower reading and math scores and increases in problem behaviors (Gibson-Davis et al. 2024). Net worth poverty also potentially affects children by affecting family functioning and contexts: it is negatively related to adult psychological and physical health and associated with higher levels of household food insecurity (Gibson-Davis et al. 2023; Keister et al. 2025).

Wealth is a related, but distinct, construct from income. Wealth is a measure of a stock of resources, whereas income is a flow; both stocks and flows are necessary for household financial stability and the absence of either is cause for concern (Killewald et al. 2017). Indeed, most children and adults who are net worth poor are not income poor, indicating that net worth poor individuals are not captured in traditional metrics of economic deprivation (Gibson-Davis et al. 2021). Across studies, the negative effects of wealth and net worth poverty are statistically independent of the effects of income and income poverty, suggesting that the absence of wealth is a risk factor in its own right (Gibson-Davis et al. 2022; Pfeffer 2018).

1.2 |. The Relative Wealth Position of Child Households

A primary reason to focus on children is that, from a sociological and life course perspective, their experiences provide insights into tomorrow’s workers, leaders, and parents. Scholars have long recognized that understanding child-rearing contexts illuminates the likely long-term health of a society (Bourdieu and Passeron 1977; Bronfenbrenner 1986; Mannheim 1952), as children’s collective experiences predict social mobility, the transmission of social and cultural capital, and political and civic trends (Lareau 2003; Pfeffer and Killewald 2018; Putnam 2000, 2015). Economic experiences during childhood can be particularly consequential. Beginning with Elder’s (1974) seminal work on the children of the Great Depression, scholars have documented that economic hardship and financial fragility experienced during childhood have lasting effects, not only on the children involved but also on the communities in which they are embedded (Chetty and Hendren 2018a, 2018b; Duncan et al. 2010; Torche 2015).

From this perspective, understanding the wealth contexts of children is important, as it likely predicts future levels of social mobility and economic health. Understanding these contexts, however, necessitates a comparison group. We may be more (or less) concerned about children’s wealth contexts if we understand whether they are unique to child households or attend to other population subgroups as well.

An understanding of wealth among child households can be clarified by a comparison with non-child and elderly households, given that many social welfare policies target these households differently. Broadly speaking, US social welfare policy prioritizes assistance according to whether individuals are “deserving” or “undeserving,” with the critical distinction being expectations regarding labor market participation (Katz 2013; Moffitt 2015). Deserving individuals are those who, because of their age or a disability, are not expected to participate in the labor market. Undeserving individuals are those who normatively should be working, traditionally defined as those who are working-age, able-bodied adults. Leaving aside considerations of disability, this classification resolves into three groups: elderly individuals, children, and working-age adults. The elderly and children are generally considered to deserve aid because their age precludes them from work (“America’s dependents,” in the words of Samuel Preston 1984), whereas working-age adults do not. Since children rely on adults, working-age households with children may also be seen as deserving. This classification—elderly, (working-age) child, and (working-age) non-child households—reflects differing economic vulnerability and shapes welfare eligibility and generosity. More aid goes to elderly households, and among working-age households, more to those with children (Moffitt 2015; Parolin, Desmond, et al. 2023).

Consistent with policies differentially targeting child, non-child, and elderly households based on their levels of economic vulnerability, we use this trifecta grouping to illuminate the relative wealth context of child households. Theoretically, child households might be disadvantaged in terms of their wealth contexts vis-à-vis elderly or non-child households for several reasons. Relative to elderly households, child households would have lower levels of wealth as predicted by the life cycle model, the typical lens used to view generational differences in net worth (Ando and Modigliani 1963). This model predicts that households with an elderly head will have greater wealth than those with a working-age head. Older adults have had more time to pay down debts and accrue assets, whereas younger adults will be less inclined to save but will instead borrow against future income. Although the life cycle model does not perfectly describe age-based patterns of wealth accumulation, its general contours, in which elderly households have higher wealth levels than non-elderly households, hold true for the United States (Keister and Moller 2000).

In addition to these age differences, several other factors suggest that levels of wealth inequality and net worth poverty may be higher for child households than for elderly households. Wealth inequality has risen in recent decades, partly because of skyrocketing wage inequality (Saez 2017). Elderly households are largely insulated against this labor market inequality because of their lower levels of labor force attachment. They may also be more likely to own their home outright than working-age heads (Goodman and Zhu 2021), allowing them to take advantage of the growth in home values without worrying about increased difficulties in buying a home. Additionally, elderly individuals have access to Social Security and Medicare. Social Security, as a steady stream of income that is indexed to the cost of inflation, may mean lower rates of net worth poverty, insofar as elderly households face less income volatility and do not have to dip into savings during times of economic crisis. Medicare (and Medicaid) insures elderly households against healthcare-related debt, which could have positive spillover effects on other areas of savings and debt.

Compared to working-age households without children, those with children likely face less favorable wealth contexts. First, modern parenting emphasizes “concerted cultivation” (Lareau 2003)—the belief that raising children requires intensive economic and temporal resources. Reflecting this, child-centered wealth activity has increased over time, with parents devoting more assets to child development (Bandelj and Grigoryeva 2021). Additionally, as described by Percheski and Gibson-Davis (2022), individuals’ financial decisions are shaped by their social identities and relationships (Granovetter 1985). Individuals engage in wealth-related activities—such as buying a home, taking out student loans, or investing in the stock market—that are consistent with their perceived identities and intimate ties (Epp and Price 2008; Zelizer 2013). Households with children therefore make decisions that reflect the presence of the child in the house, and such decisions may lead to lower levels of wealth. For example, households with children may make wealth-depleting choices, such as buying a home in a costly area for better schools. They may also make labor market decisions that affect savings and debt, like accepting lower pay in exchange for health insurance. To be clear, differences in wealth behaviors are likely sharper between parents and nonparents than between households with and without a child present, given that some households without a child present will have children residing elsewhere. Nevertheless, having a child in the household likely shifts wealth behaviors relative to households without a resident child.

Limited research has explored how, in the aggregate, child and non-child households differ in wealth contexts. Most of the work in this area has instead focused on how the transition to parenthood affects the wealth levels of individual households. Generally, residing with children is associated with declines in wealth, although the magnitude of the effect varies widely (Maroto 2018; Maroto and Aylsworth 2017; Painter and Shafer 2011). Children who live in a single-parent home with a female household head might have the largest negative effects on household wealth (Percheski and Gibson-Davis 2022). We are not aware of any study that has compared wealth trends over time for child and non-child households.

1.3 |. Wealth Trends and COVID-Era Spending

From the 1980s to the late 2010s, a period of rising wealth inequality emerged primarily through increases in wealth for those at the top (Wolff 2016, 2022). Wealth inequality began to climb in the 1980s before plateauing somewhat in the 1990s and early 2000s. The Great Recession (2007–2010) led to wealth losses across the population, but it also increased wealth inequality because the wealthiest individuals lost proportionally less (Pfeffer et al. 2013). Since the Great Recession ended, median household levels have increased, but the top of the wealth distribution has seen disproportionately larger gains (Aladangady et al. 2023). During the 2010s, wealth levels among households in the bottom 50% were either stagnant or declining, contributing to higher wealth inequality levels in 2019 than in the previous four decades (Wolff 2017, 2022).

Against this period of rising wealth inequality, the early 2020s merit special attention because of the large increases in government spending that occurred in response to the COVID-19 pandemic. Much of this increase was directed toward households with children. On a per capita basis, per child expenditures in 2021 were $10,720—$4000 higher than 2019 expenditures. Economic stimulus payments that were provided to all US households included an additional $500 for each qualifying resident child (U.S. Congress 2020). The government also increased the generosity levels of both the Child Tax Credit and the Child and Dependent Care credit; the Child and Dependent Care credit was also made fully refundable (U.S. Congress 2021). This child-centered response to the pandemic-related economic crisis stands out for its magnitude; spending on children in 2021 was 2.9% of GDP, higher than the Great Recession-driven peak of 2.5% in 2010 (Lou et al. 2022).

Increased government spending had positive economic impacts on households with children, particularly those that were lower resourced (Parolin, Ananat, et al. 2023; Pilkauskas et al. 2022; Wimer et al. 2022). Among families in the bottom 20% of the income distribution, mean income increased by nearly $1000 between 2020 and 2021, reaching a level ($22,120) close to pre-pandemic levels (U.S. Census Bureau n.d.). The Supplemental Poverty Measure (SPM), which accounts for taxes and transfers (among other factors), was 5.2% for children in 2021, the lowest child SPM rate ever recorded (Creamer et al. 2022). The 45% decline in the SPM between 2020 and 2021 was also the largest year-over-year decline since the federal government began keeping SPM statistics (Creamer et al. 2022). Child poverty rates based on the Official Poverty Measure (OPM), which do not measure stimulus payments or the CTC, also declined (albeit more modestly) from 16.0% in 2020 to 15.3% in 2021. However, child poverty rates did not remain depressed for long: by 2022, the SPM for children rose to 12.4% (Shrider and Creamer 2023). Nevertheless, the decline in poverty rates demonstrates how government expenditures can affect the financial health of lower-resourced families.

Whether such spending had similar impacts on wealth and wealth deprivation among child households is an open question. Theoretically, impacts on wealth might seem unlikely, given that lower-income households have less of an economic buffer than higher-income households and might respond to cash transfers by increasing consumption rather than changing savings or debt levels. However, the government expenditures that were common during the pandemic—lump-sum payments through stimulus checks or via the tax system—might favor wealth-building (Jones and Michelmore 2018; Parolin, Ananat, et al. 2023; Shaefer et al. 2013). Variation in the CTC distribution during the pandemic illustrates this point: it was distributed as a monthly cash benefit from July to December 2021 and as a lump-sum payment through a tax rebate in 2022. The monthly payments were used to buy food and other basic necessities, whereas the lump-sum payment was used to pay down housing debt and accrue savings (Parolin, Ananat, et al. 2023). Therefore, it seems likely that, akin to the effects on income poverty, these transfers decreased net worth poverty.

1.4 |. Hypotheses

This study uses 35 years of data, covering 1989–2022, to compare wealth trends and net worth poverty levels among three household types. Based on previous work (Gibson-Davis and Percheski 2018), we hypothesize that child households will compare unfavorably with elderly households in terms of wealth inequality levels, wealth levels, and net worth poverty rates. Further, we expect that wealth disparities between child and non-child households and between child and elderly households have grown over time, with one exception: the early 2020s. The provision of government aid during the early 2020s might have coincided with increases in wealth for child households, narrowing disparities between child households and both non-child and elderly households.

2 |. Data and Methods

Data are from the 1989 to 2022 waves of the Survey of Consumer Finances (SCF), a repeated cross-sectional study of US households conducted by the Federal Reserve every 3 years. The SCF is considered the premier source of US wealth data (Aladangady et al. 2023). It is the best data source for this project because it is the only dataset that captures the full range of the wealth distribution, an important consideration given the extreme right skew of US wealth (Keister 2014; Kennickell 2008). It is also the only large dataset that has wealth information for 2022, permitting an analysis of wealth after the COVID-19-related expenditures were fully implemented. The SCF’s unit of analysis is the primary economic unit, which is akin to the US Census Bureau’s definition of a household. We therefore refer to our unit of analysis as households. To maximize our observation period, we used data from the earliest wave available (pre-1989 waves are not compatible with later years). Our 34-year observation period spans four decades and can be used to understand any changes in the appropriate context. The total sample size is 58,148 households.

Households in our sample were stratified into three mutually exclusive household types: child households (with at least one resident member under age 18 and no member aged 65 or older; n = 19,877), non-child households (with no household member younger than 18 or aged 65 or older; n = 24,603), and elderly households (with at least one household member aged 65 or older; n = 13,668). A small fraction of elderly households (n = 522) included a child under age 18. Classifying these 522 households as child households, however, did not change the results. Note that non-child households may have a child residing elsewhere; data limitations precluded us from identifying households whose children live outside the home.

Our key economic variables are wealth, net worth, poverty, income, and income poverty. We define wealth as the sum total of a family’s fungible assets minus debts. The SCF measures the presence and value of 12 categories of assets: savings and checking accounts, certificates of deposit, pooled investment accounts, stocks, bonds, retirement accounts, vehicles, future pensions not yet realized, primary residence and other real estate, business assets, tangible assets (i.e., art and jewelry), and assets not classified elsewhere. Seven categories of debts are assessed in the SCF: mortgages on the primary residence, mortgages on other real estate, business debt, credit card debt, educational debt, vehicle loans, and other liabilities. (Medical debt is not specifically assessed but is classified into installment debt or the “other debt” category, depending on how it was paid.) Following Wolff (2016), we excluded the value of vehicles (whose resale value is far less than the consumption value) and the value of future pension and Social Security income (income that the household might eventually realize but cannot yet readily access). By including only assets that are readily convertible to cash, our measure of net worth reflects a household’s current holdings. A supplementary analysis indicated that including vehicles and future pensions in the definition of wealth did not substantively alter our findings. Wealth was converted to 2022 dollars using the Consumer Price Index.

To account for economies of scale, we adjusted our net worth variable by the square root of household size. (Results, however, were largely the same when we did not adjust for household size). On average, child households had 2.1 and 2.2 more members than non-child and elderly households, respectively. The number of child household members remained virtually constant over our study period, indicating that changes in household size do not explain results.

Net worth poverty is based on the continuous measure of wealth. Households are net worth poor if their net worth is less than one-quarter of the federal poverty line, adjusted for family composition (Gibson-Davis et al. 2021; Haveman and Wolff 2004). Net worth poverty thus represents having sufficient resources for a 3-month period. Although arbitrary, 3 months’ worth of resources may correspond to the savings displaced when households suffer an income shock (see Brandolini et al. 2010 for more on this issue). In 2022, a household with two adults and two children was considered income poor if total household income was less than $29,678 (U.S. Census Bureau 2023b) and would be considered net worth poor if their net worth was less than $7,420.

The SCF assesses income as a before-tax measure that sums wage and business income; dividends, realized capital gains, and other stock-related income; pensions and Social Security income; select government transfers, such as unemployment and the Supplemental Nutrition Assistance Program (SNAP); alimony and child support; and income sources not otherwise mentioned. The SCF did not collect specific information on subsidy payments or tax credits. Income, like wealth, was converted to constant 2022 dollars. We used this continuous measure of income to classify households as income poor, again using the federal poverty line for thresholds. Note that because the SCF counts income sources that are excluded from the official definition of poverty (e.g., SNAP), the SCF poverty measure will not accord with the OPM.

To supplement the SCF income poverty measure, we also use the SPM, which is a more comprehensive measure of income sufficiency than the OPM because it counts both cash and noncash income sources and subtracts expenses and taxes. Program benefits counted in the SPM include income from tax programs (e.g., the Earned Income Tax Credit), food-related benefits (SNAP; school lunch programs; and Women, Infants, and Children), and housing and energy subsidies. Deductions include state and federal taxes, as well as work, childcare, and medical expenses. The SPM threshold also adjusts for geographic differences in the cost of living. Yearly SPM estimates are from Wimer et al. (2023), given that the SCF does not collect SPM data.

Using these data on wealth, net worth, poverty, income, and income poverty, we provide descriptive statistics on how levels of wealth and income and poverty rates changed over the observation period. Inequality measures include the Gini coefficient, which summarizes levels of inequality with one estimate and can readily be compared across populations. The Gini coefficient is measured on a scale of 0–1, where 0 = complete equality and 1 = complete inequality (a one-unit change in the Gini coefficient is not readily interpretable, except to indicate higher inequality). We also present median wealth levels for households at relative points in the wealth distribution (the top 1%, the next 9%, and so on). Wealth distributions were calculated separately for each household type. Estimating household median wealth at different points in the wealth distribution is akin to estimating median wealth at the midpoint of that distribution. For example, household median wealth in the bottom 50% is the average of the median wealth for households at the 25th and 26th percentiles. We also compared changes in net worth and income poverty, both by household type and across time, to understand the relative disparities between the groups on these poverty measures as well as to understand trends over time. All analyses were weighted. Note that because we used constant dollars, we adjusted for overtime variation in wealth, income, and poverty thresholds.

3 |. Results

3.1 |. Trends in Income and Wealth Inequality

Our first analysis describes trends in wealth inequality from 1989 to 2022 for child, elderly, and non-child households, as measured by the Gini coefficient (Figure 1). For completeness, we also include trends in income inequality. As prior research would suggest (Aladangady et al. 2023), within each group, income inequality was lower than wealth inequality. Further, trends in income inequality did not necessarily track with those of wealth inequality, demonstrating the difference between these two measures of inequality.

FIGURE 1 |.

Gini coefficient, wealth and income, by household type, 1989–2022. Panel A: Child HHs. Panel B: Non-child HHs. Panel C: Elderly.

Across periods, child households had lower levels of wealth inequality than elderly households but had comparable levels to non-child households. Child and non-child households followed a similar pattern: small increases in levels of inequality in the 1990s and 2000s, followed by a sharp increase in the early 2010s, corresponding to the Great Recession. Wealth inequality peaked in non-child households in 2016 at 0.90 before falling to 0.87 by 2022. The peak in wealth inequality was slightly higher and earlier for child households, reaching 0.93 in 2010 and declining thereafter. Levels of wealth inequality for elderly households were lower than those for the other household types throughout the observation period, but they did rise slightly from 0.78 in 1989 to 0.82 in 2022.

Wealth inequality levels for child households are notable in one respect in terms of the scale of the decline observed between 2019 and 2022. The 6-percentage point decline is the largest observed between any two time points, for both income and wealth inequality, and for any household type. It also mirrors the 6-percentage point increase in wealth inequality for child households between 2007 and 2010. Because of this decrease in wealth inequality, child households are the only household type whose levels of wealth inequality in 2022 (0.84) were comparable to those in 1989 (0.83). Additionally, this decrease resulted in a narrowing of the wealth disparities between elderly households and child households. In 1989, the difference in the Gini coefficient between elderly and child households was 5 percentage points; by 2022, it was reduced by half to 2.5 percentage points.

3.2 |. Changes in the Wealth Distribution

Our next set of estimates presents changes in median wealth for households at relative points in the wealth distribution (Table 1). We examined changes over select periods to reflect years of economic growth or retraction, concentrating on 2007–2022, when changes in wealth were most evident for child and non-child households.

TABLE 1 |.

Median net worth for selected points in the wealth distribution, household types, by year.

	Child households
	P0-P100	P0-P50	P51-P90	P91-P99	P100
1989	35,544	161	94,359	485,946	2,617,700
1995	28,743	181	75,931	485,487	2,882,562
2007	41,829	0	144,201	847,421	6,778,104
2010	16,747	−418	70,811	852,249	6,637,688
2013	13,688	−514	82,773	758,317	6,432,183
2016	21,839	0	99,475	874,709	9,093,957
2019	36,721	114	117,414	1,087,289	9,661,466
2022	62,621	2178	172,364	1,300,104	7,099,922
Percent change between:
1989–2022	76.2	1254	82.7	167.5	171.2
1989–2007	17.7	−100.0	52.8	74.4	158.9
2007–2010	−60.0	—	−50.9	0.6	−2.1
2010–2013	−18.3	−23.0	16.9	−11.0	−3.1
2013–2016	59.6	−100	20.2	15.3	41.4
2016–2019	68.1	—	18.0	24.3	6.2
2019–2022	70.5	1,814	46.8	19.6	−26.5
	Non-child households
	P0-P100	P0-P50	P51-P90	P91-P99	P100
1989	55,532	1169	177,649	1,128,654	5,851,643
1995	54,048	1588	164,108	840,080	6,921,897
2007	91,956	2867	289,339	1,597,755	13,200,000
2010	50,043	102	194,681	1,612,892	8,875,584
2013	41,065	3	179,037	1,234,643	9,536,969
2016	49,879	83	169,068	1,555,021	13,400,000
2019	50,897	339	178,669	1,511,381	11,900,000
2022	74,790	604	249,753	1,836,099	11,700,000
Percent change between:
1989–2022	34.7	−48.3	40.6	62.7	99.9
1989–2007	65.6	145.3	62.9	41.6	125.6
2007–2010	−45.6	−96.4	−32.7	0.9	−32.8
2010–2013	−17.9	−97.5	−8.0	−23.5	7.5
2013–2016	21.5	3147	−5.6	25.9	40.5
2016–2019	2.0	308	5.7	−2.8	−11.2
2019–2022	46.9	78	39.8	21.5	−1.7
	Elderly households
	P0-P100	P0-P50	P51-P90	P91-P99	P100
1989	131,007	31,715	281,203	1,486,177	8,479,260
1995	156,905	42,834	291,903	1254,058	9,847,455
2007	237,853	73,080	522,294	2,842,233	14,100,000
2010	208,937	55,492	432,365	2,469,752	15,200,000
2013	186,089	52,225	397,182	2,547,348	13,900,000
2016	221,078	48,687	511,169	3,186,616	19,400,000
2019	221,497	61,060	478,099	3,060,315	20,800,000
2022	278,626	69,206	623,213	4,355,842	25,700,000
Percent change between:
1989–2022	112.7	118.2	121.6	193.1	203.1
1989–2007	81.6	130.4	85.7	91.2	66.3
2007–2010	−12.2	−24.1	−17.2	−13.1	7.8
2010–2013	−10.9	−5.9	−8.1	3.1	−8.6
2013–2016	18.8	−6.8	28.7	25.1	39.6
2016–2019	0.2	25.4	−6.5	−4.0	7.2
2019–2022	25.8	13.3	30.4	42.3	23.6

Note: Net worth amounts are in constant 2022 dollars. Net worth was adjusted by the square root of household size. All estimates are weighted. Sample sizes: Child households: n = 19,877; non-child households: n = 24,603; elderly households: n = 13,668. Year sample sizes 1989: n = 3143; 1995: n = 4299; 2007: n = 4417; 2010: n = 6482; 2013: n = 6015; 2016: n = 6248; 2019: n = 5777; and 2022: n = 4595.

Although the period saw wealth growth for both child and elderly households, child household wealth levels were lower than those of elderly households in every year and at every point in the distribution. Disparities between elderly and child households were smallest among the wealthiest households and largest among the least wealthy. In 2022, among the top 1% of households, elderly households had wealth levels that were more than 3.5 times as high as those of child households ($25,700,000 vs. $7,099,922). Among the bottom 50%, however, elderly households had wealth levels that were more than 30 times as high ($69,206 vs. $2,178). With the exception of 2019–2022, disparities between elderly and child households grew over time. In 1989, at the median (e.g., the “P0-P100” column), child households had wealth levels that were 27% of those of elderly households ($35,544 vs. $131,007). By 2019, child households had wealth levels that were 17% of those of elderly households. For those in the bottom 50%, wealth levels for child households were 0.5% of elderly households in 1989 and 0.2% in 2019.

The wealth gap between child and elderly households narrowed between 2019 and 2022, the only period when disparities decreased. Between 2019 and 2022, wealth levels rose by 70.5% for all child households and by 1,814% among those in the bottom 50%. Wealth levels also increased for elderly households, but by a smaller amount (26% for all households and 13.3% for the bottom 50%). As a result, although elderly households still had substantially higher levels of wealth than child households in 2022, their relative wealth advantage narrowed, particularly among households in the bottom 50%. In 2019, the child-t o-e lderly wealth ratio among those in the bottom 50% was 0.002:1, indicating that child households had 0.002 cents for every dollar of elderly wealth. By 2022, that ratio had narrowed to 0.03, the smallest difference observed at any time point.

The wealth contexts of child households also compared unfavorably to non-child households. Across years and points in the distribution, child households had lower levels of wealth than non-child households (the exception was 2022, as we discuss later). Differences between non-child and child households were smaller than those observed between elderly and child households. Nevertheless, median wealth levels were substantially larger among non-child households than among child households. In 2019, at $50,897, non-child median wealth levels were 38% higher than those of child households. As we found when comparing elderly and child household wealth, disparities between child and non-child households were largest for those in the bottom 50%. In 1989, child households had 14 cents of wealth for every dollar of wealth for non-child households; by 2016, this ratio had fallen to less than 1 cent. Again, however, the 2019–2022 period was an exception, reversing the trend of increasing disparities but only among households in the bottom 50%. In 2019, among those in the lower half of the distribution, child households had wealth levels that were half as high as those of non-child households ($114 vs. $339). By 2022, that ratio had flipped, with child households having wealth levels ($2,178) that were 3.6 times as high as those of non-child households ($604).

Results thus far indicate that child households have lower levels of wealth than elderly or non-child households, with the largest differences observed among the least wealthy households. This pattern held throughout the observation period, with the exception of the early 2020s. Because of the relatively large increases observed for child households between 2019 and 2022, the wealth disparities among those in the bottom 50% narrowed between child and elderly households and reversed for child and non-child households. In 1989, among those in the bottom 50%, child households had half a cent ($0.005) of wealth for every dollar of elderly wealth; by 2022, that ratio was 3 cents ($0.03). Comparative figures for the ratio of child to non-child households were $0.13 in 1989 but $3.64 in 2022.

3.3 |. Changes in Net Worth Poverty and Income Poverty

We next compare net worth poverty levels across household types. We begin by presenting results for changes in net worth poverty (Figure 2) and then compare changes in net worth poverty with those of income poverty (Table 2). To increase readability, we present results for select survey waves in Figure 2. Table 2 includes results for all survey waves.

FIGURE 2 |.

Net worth poverty, by household type, 1989–2022. Sample sizes: Child households: n = 19,877; non-child households: n = 24,603; and elderly households: n = 13,668.

TABLE 2 |.

Poverty measures, by household type.

	Child households			Non-child households			Elderly households
	NWP	SCF-IP	SPM	NWP	SCF-IP	SPM	NWP	SCF-IP	SPM
1989	31.8	16.3	17.6	28.8	8.1	9.4	13.5	11.0	14.3
1992	33.5	18.6	19.7	27.0	9.1	10.4	11.6	12.4	16.1
1995	31.9	17.9	16.9	27.0	9.9	11.0	11.4	10.6	14.7
1998	31.7	16.1	15.2	28.1	9.3	10.0	10.7	10.5	13.9
2001	30.2	12.5	12.7	26.8	10.0	11.5	10.7	10.6	15.2
2004	30.7	13.7	12.2	26.8	10.9	12.3	11.2	10.5	15.6
2007	33.5	13.8	13.4	25.5	9.8	12.2	11.1	10.0	16.9
2010	40.1	17.3	14.8	31.9	12.2	14.7	13.0	10.1	16.2
2013	41.8	16.6	14.6	34.8	13.2	16.0	11.6	10.6	15.8
2016	37.4	14.8	12.8	33.6	11.8	13.9	13.8	10.1	15.0
2019	35.4	11.5	10.1	31.9	9.8	11.9	13.4	9.1	12.8
2022	26.5	12.4	10.3	28.6	10.8	12.5	13.7	10.3	13.8

Note: NWP: net worth poor, based on Survey of Consumer Finances. SCF-IP: income poor, based on the Survey of Consumer Finances. SPM: income poor, based on the Supplemental Poverty Measure. Sample sizes for Survey of Consumer Finances: Child households: n = 19,877; non-child households: n = 24,603; elderly households: n = 13,668. Year sample sizes 1989: n = 3,143; 1995: n = 4,299; 1998: n = 4,305; 2001: n = 4,519; 2004: n = 4,519; 2007: n = 4,417; 2010: n = 6,482; 2013: n = 6,015; 2016: n = 6,248; 2019: n = 5,777; and 2022: n = 4,595.

Across time points, child households had substantially higher net worth poverty rates than elderly households. In 2022, for example, the net worth poverty rate for child households was 26.5%, almost twice as high as the net worth poverty rate among elderly households (13.7%). The size of the disparity in net worth poverty rates varied over time, driven by variation in child households (elderly poverty rates were relatively flat, varying by less than 3 percentage points between 1989 and 2019). Child net worth poverty rates increased substantially with the onset of the Great Recession, rising from 33.5% in 2007 to a peak of 42% in 2013. They then began to decrease, with a 9-percentage point decline between 2019 and 2022. Because of this decline, the net worth poverty gap between elderly and child households in 2022 (13 percentage points) was the smallest of any observed period and was half the size of the 2013 gap (30 percentage points).

Child households generally had higher net worth poverty rates than non-child households. Non-child net worth poverty rates ranged from 25.5% (2007) to 34.8% (2013), roughly 38 percentage points higher than rates for child households. Like child households, non-child households had their lowest net worth poverty rates in the years before the Great Recession, with rates rising with the recession’s onset and then falling as it receded. And though net worth poverty rates fell for non-child households between 2019 and 2022, the size of the decline was smaller in both relative and absolute terms than the same decline observed for child households. As a result, net worth poverty rates in 2022 were higher for non-child households (28.6%) than for child households (26.5%), the only time point at which that occurred.

A comparison of income and net worth poverty rates by household type and across time (Table 2) can illustrate how these groups compare on these measures and whether disparities in net worth poverty rates between groups exceed disparities seen in income poverty rates. Disparities in net worth poverty rates between elderly and child households are larger than those seen for income poverty rates. Relative to child households, elderly households had lower income poverty rates throughout the period using the conventional measure of income poverty (e.g., the SCF income poverty measure) but slightly higher rates from 2001 onward when using the SPM. Differences in income poverty rates between the two groups were relatively small, between 2 and 7 percentage points. In contrast, the net worth poverty rate difference between elderly and child households was substantially larger: leaving 2022 (with a gap of 13 percentage points) aside, the disparity in net worth poverty rates ranged between 18 and 30 percentage points. The wide disparity in net worth poverty rates, which is not mimicked by disparities in income poverty rates, suggests that child households may be disadvantaged vis-à-vis elderly households in a way that has not been previously recognized.

The comparison between child and non-child households suggests that these two groups are more comparable in their income and net worth poverty rates, with smaller differences in the income poverty and net worth poverty gaps. Nevertheless, after 2000, the relative gap in net worth poverty rates exceeded the relative gap in income poverty rates. Before 1998, income poverty rates were 5–9 percentage points higher among child households than among non-child households, exceeding the same-year gap in net worth poverty. After 2000, however, the relative gap in income poverty rates between child and non-child households contracted, whereas the gap in net worth poverty rates increased. However, the differences in net worth poverty rates between the two groups were relatively small, ranging from 3 to 8 percentage points.

This overtime comparison of income poverty rates and net worth poverty rates provides useful context for understanding the decline in net worth poverty rates seen for child households between 2019 and 2022. Overall, net worth poverty rates and income poverty rates followed a similar pattern for child and non-child households: they changed relatively little before 2007, increased with the onset of the Great Recession, and declined from the early to mid-2 010s. Elderly households had more consistent income and net worth poverty rates, and they were seemingly less affected by the economic disruption of the Great Recession or the economic recovery that followed. Against these trends, the 2019–2022 change in net worth poverty for child households stands out. In both absolute and percentage terms, this change (9 percentage points, or a 33% decline) was the largest 3-year shift, either positive or negative, observed for net worth poverty or income poverty for any of the three household types. It was roughly one-third larger than the next-largest shift, when the net worth poverty rate increased by 6.9 percentage points between the Great Recession in 2007 and 2010.

Notably, the net 2019–2022 decline in net worth poverty for child households was not mimicked by a decline in income poverty, as measured through the SCF or the SPM. Income poverty rates for child households rose between 2019 and 2022, albeit modestly (e.g., from 10.1 to 10.3 for the SPM). Thus, while net child household worth poverty rates were exhibiting the largest 3-year decline in at least three decades, child household income poverty rates were moving in the opposite direction. (For consistency, we show SPM estimates only for the years of SCF data collection. However, the SPM estimates in the early 2020s show the expected pattern, falling between 2020 and 2021 before rising again in 2022.) Interestingly, 2019–2022 was the only period when net worth and income poverty moved in opposite directions.

In supplementary analyses, we considered whether housing could have accounted for the 2020–2022 wealth increases for child households. Housing is the most important asset for child households (Gibson-Davis and Hill 2021), and home values rose nearly linearly between 2020 and 2022 (St. Louis Fed n.d.). We found that even after housing equity was excluded, median wealth levels for child households increased between 2019 and 2022, except among the top 1%. Additionally, net worth poverty decreased between 2019 and 2022, albeit more modestly (5 percentage points) than when housing equity was included. These results suggest that the increase observed between 2019 and 2022 was not driven entirely by housing.

In sum, child households generally had higher net worth poverty rates than non-child and elderly households. Child and elderly households had more comparable income poverty rates than net worth poverty rates, but the relative gaps among child and non-child households were more muted. Higher poverty rates held for all years except 2022, when net worth poverty rates for child households fell below those of non-child households. The narrowing of this disparity reflects the steep decline in net worth poverty rates for child households seen between 2019 and 2022. These declines in net worth poverty for child households were not mirrored by changes in income poverty; indeed, income poverty for child households increased over the same period.

4 |. Conclusion

To advance knowledge about whether wealth vulnerabilities are specific to child households, this study provides the first comparison of overtime trends in wealth inequality, wealth levels, and net worth poverty rates between child, non-child, and elderly households. This comparison illustrates whether the wealth profile of child households—with high inequality rates, low wealth levels, and high net worth poverty rates—is specific to that subpopulation or reflects the US wealth context more generally.

Consistent with our hypothesis, elderly households have a more favorable wealth profile than child households. In line with previous work (Gibson-Davis and Percheski 2018), child households had higher levels of wealth inequality and lower levels of wealth than elderly households. Wealth differences were evident throughout the distribution, but they were largest among the least wealthy households. In 2019, for example, among those in the bottom 50% of households, child households had 0.002 cents worth of wealth for every dollar of wealth held by an elderly household. Child households also had substantially higher net worth poverty rates than elderly households. The net worth poverty rate for elderly households varied between 11% and 13%, roughly 20–30 percentage points higher than that of child households (with the exception of disparities in 2022, a point we will return to). The child and elderly disparity in net worth poverty rate far exceeds disparities seen for income poverty. These findings suggest that differences between “America’s dependents” (Preston 1984) are largest when the least wealthy households are considered and are perhaps larger than when income poverty is the only metric of economic deprivation used. These findings reinforce Preston’s larger concern that among these economically vulnerable populations, children remain particularly disadvantaged.

We also hypothesized that child households would have worse wealth contexts than non-child households, in part because the social identity of parenthood may lead individuals to make child-centric choices that could lower household wealth. Results were largely consistent with this hypothesis. Leaving 2022 aside, non-child households consistently had higher levels of wealth and lower levels of wealth inequality than child households. Differences were largest for households among the bottom 50% of the distribution, where child households’ wealth levels were pennies on the dollar relative to non-child households. In addition, net worth poverty rates were 3–8 percentage points lower for non-child households than for child households, depending on the year. After 2000, differences in net worth poverty rates also exceeded those for income poverty rates, although the scale of the differences was small. Differences between child and non-child households were less pronounced than those between child and elderly households. Nevertheless, results suggest that the wealth contexts of child households are generally worse than those of non-child households.

Our final hypothesis was that differences between child and elderly households and between child and non-child households would generally grow over time but might have attenuated in the early 2020s, coinciding with larger government expenditures in response to the COVID-19 pandemic. Our results largely support this hypothesis. Wealth disparities between child households and the other two household types generally grew over the 1990s and 2000s and accelerated with the onset of the Great Recession. In the mid- to late 2010s, however, the wealth position of child households began to improve, slightly narrowing the wealth disparity with the other two household types. Changes in wealth favoring child households accelerated between 2019 and 2022—a period of unprecedented increases in wealth among child households, particularly among those in the bottom half of the wealth distribution. Among this group, median wealth levels rose by nearly 2,000% between 2019 and 2022. Additionally, the 2019–2022 period corresponded with the largest observed change in net worth poverty (by 9 percentage points, representing a 33% decrease). These changes in net worth poverty also far exceeded any changes seen for income poverty. Future work should investigate whether increases in government expenditures caused the observed changes in wealth and net worth poverty levels. If a causal relationship can be established, then policymakers would have a known tool at their disposal to improve the wealth contexts of child households.

Limitations to this study should be noted. This descriptive study cannot indicate a causal link between policy changes and changes in wealth as other factors related to the pandemic may have also played a role. For example, households may have increased their savings because of fewer opportunities to spend on leisure or vacation, or they may have spent relatively less on debt because of rent forgiveness or pauses in student loan repayments. Housing prices also increased (Federal Reserve Bank of St. Louis 2025), leading to increases in assets (although supplementary analyses suggest that the observed increase was robust to changes in housing equity). However, data limitations preclude identifying compositional shifts in asset and debt patterns. We are not able to observe whether assets increased because of stimulus payments or whether debts decreased because some loan obligations were put on hold. This aggregate analysis may also have overlooked important within-child household variation. Some household types, such as those headed by a single parent, may have been at an economic disadvantage during the COVID-19 pandemic because of labor market fragility or increased caregiving responsibilities. Additional work is needed to understand why wealth levels changed during this period, why these changes were focused among the least wealthy child households, and whether these wealth gains varied by demographic type.

Relatedly, the household types under consideration are broad and demographically varied, both within and across groups, in ways that likely affect their wealth. Nevertheless, the approach used here to divide households by the head’s age and the presence of a child forms a useful trifecta for thinking about the context of US wealth and whether the wealth context of child households is unique. Finally, because our analyses extend only through 2022, we do not know whether decreases in wealth inequality, increases in wealth, and decreases in net worth poverty for child households will continue. Analyses of child income poverty rates indicated that they rose once COVID-era spending was curtailed (Shrider and Creamer 2023), and net worth poverty may likewise have increased. However, insofar as net worth poverty is a cumulative measure of assets and debts (as opposed to income poverty, which reflects the absence of a flow of resources), net worth poverty changes might be more enduring. Still, we can only speculate about how wealth levels and net worth poverty rates may change in the future.

This study investigated the uniqueness of the wealth context of child households by comparing it with that of elderly and non-child households. Findings indicate that the trends that characterize US wealth more generally—high inequality rates with low wealth levels for those at the bottom of the distribution—are particularly acute for child households. Given that wealth influences children’s life chances and informs future levels of wealth mobility, these findings are not reassuring. An important exception to this general pattern of findings was the early 2020s, when disparities in wealth levels and net worth poverty rates either narrowed (between elderly and child households) or reversed direction (between child and non-child households). This time frame coincided with a substantial uptick in government expenditures directed toward children; future work will need to investigate whether this correlational association is causal. At the very least, however, these results suggest that inequality and wealth among child households are not immutable and can decrease over time. Given that wealth promotes child flourishing (Gibson-Davis et al. 2022; Miller et al. 2021; Pfeffer 2018), increasing wealth in the early 2020s for child households, particularly for those in the bottom 50%, is heartening.

Data Availability Statement

Survey of Consumer Finances data can be obtained through their website and is available to all (https://www.federalreserve.gov/econres/scfindex.htm).