6
Children’s Family Income, Wealth, and Parental Employment
Family economic resources and parental employment during childhood play an important role in shaping the family, schooling, and neighborhood contexts where children develop. These contexts, in turn, are key for children’s academic skills, health, and opportunities to secure good jobs when they become adults. Persistently low levels of family income, wealth, and employment may therefore be important drivers of intergenerational poverty, while policies that increase parental incomes, wealth, and employment may increase intergenerational mobility.
In assessing the role of family economic factors as a driver of intergenerational poverty, we begin with data on trends in U.S. incomes and wages and comparisons of the extent of child poverty in the United States and other industrialized Anglophone countries. Employment rates and earnings for lower-skilled workers in general—and for the women and people of color in this group in particular—remain lower than for other populations, despite wage growth for the bottom quintile in the late 2010s and early 2020s and supplementation from the work-based safety net. Moreover, gaps between workers and families at the bottom (and middle) of the income distribution and workers and families in the top tier grew substantially.
We then provide a brief review of the literature on the correlational and causal connections between children’s household income, parental employment, and well-being in adulthood. With regard to household income, we point out that the introduction of safety net programs such as the Supplemental Nutrition Assistance Program (SNAP; formerly known as Food Stamps) and more recent changes in the Earned Income Tax Credit (EITC) appear to have reduced intergenerational poverty. However, the
intergenerational impacts of further expanding safety-net benefits are less certain, especially if doing so does not also increase parental employment. Furthermore, rigorous evidence suggests that increased parental employment improves child outcomes only when it also raises family incomes. Evidence on the causal impact of wealth on intergenerational poverty is suggestive, but does not support strong conclusions.
Given direct evidence linking reductions in intergenerational poverty to programs that increase both family income and parental employment, the committee outlines several possible expansions of the EITC (perhaps accompanied by other changes in tax credits) that appear to be most likely to reduce intergenerational poverty. We then list a number of other program and policy ideas that are supported by indirect evidence.
TRENDS IN INCOME AND EARNINGS
Family Income and Child Poverty
To determine the extent to which family income is associated with intergenerational poverty, it is important to examine how intergenerational mobility is affected by both the absolute levels of family income and earnings and the degree of income and earnings inequality between low- and higher-skilled parents. Poverty persistence across generations would seem to relate most closely to the absolute levels of children’s family income, in both childhood and adulthood. However, as reviewed in Chapter 2, evidence on trends in intergenerational mobility presented in Chetty et al. (2017) shows that while trends in national income play an important role, an even more significant factor is how the gains from economic growth are shared between families at the top and bottom of the income distribution. Accordingly, we will focus on the absolute levels of income and earnings as well as on inequality.
Trends in children’s family income can best be illustrated by showing the evolution of the average household incomes of children whose families are in the bottom, middle, and top of the income distribution. Figure 6-1 provides this information for 1967–2019, using data from the U.S. Census Bureau. The income data include both the usual sources of cash income, such as earnings, and in-kind income sources including tax credits and payments from programs such as SNAP.1 The average annual household
___________________
1 Dollars of cash and in-kind income received in different years are adjusted for inflation using the Personal Consumption Expenditures Price Index (PCEPI). Inflation measured by the PCEPI over this period is more modest than when measured using another popular inflation adjustment, the CPI-U-RS, and thus inflation-adjusted measures increased more (or fell by less) than they would if we used the CPI-U-RS. Another possible adjustment for these kinds of income comparisons is for family sizes, which were considerably larger in the 1960s and 1970s than they are today. As noted in Chapter 5, cash and in-kind sources of transfer income, and the programs that provide them, may affect total household resources in the same ways.
NOTES: Figure 6-1 shows the average household incomes for children in the bottom, middle, and top income quintile. Household income includes all post-tax cash income and in-kind resources counted under the Supplemental Poverty Measure with the exception of nondiscretionary expenses. Adjustment to $2021 are made using the PCEPI. Due to higher rates of nonresponse to survey questions about income and resources at the top and bottom end of the income distribution, and to changes in the U.S. Census Bureau’s top-coding procedure across time (Bollinger et al., 2018), we do not include households in the top or bottom 2.5% of the distribution when producing within-quintile averages in the top and bottom quintiles, respectively. See Appendix C: Chapter 6 for a discussion of the sensitivity of the results to this decision.
SOURCE: Data for 1967–2009 produced with the Historical Supplemental Poverty Measure Data series (Wimer et al., 2022) and the Annual Social and Economic Supplement to the Current Population Survey (CPS-ASEC) data from IPUMS-CPS (2022). Results for 2010 to 2020 produced with the CPS-ASEC data from IPUMS-CPS (2022). Income adjusted to $2021 using the PCEPI, produced by the U.S. Bureau of Economic Analysis and retrieved from FRED, Federal Reserve Bank of St. Louis.
incomes of children in the bottom quintile increased during this period by 80%, from about $20,000 to $36,000. For children in the middle of the income distribution, the increase amounted to more than $40,000, but the percentage change (+90%) was only a little larger than the increase for low-income children. Both absolute (+$116,000) and relative (+146%) household income increases were largest for children in families at the top of the distribution.
The near doubling of household incomes at the bottom end of the distribution over the past half century has significantly reduced child poverty. Using an Supplemental Poverty Measure (SPM)-based poverty measure, National Academies (2019a; Figure S-2) found that child poverty rates fell by 45% between 1967 and 2016. The U.S. Census Bureau documents further declines through 2020, although it still counted 9.7% of children as living in poverty in 2020 (Shrider and Creamer, 2023). Pandemic programs such as the expansion of the Child Tax Credit (CTC) produced a dramatic further reduction in the nation’s child poverty rate—from 9.7% to 5.2%—between 2020 and 2021. However, the CTC expansion and other income-based pandemic programs ended in December 2021 and the SPM-based child poverty more than doubled to 12.4% in 2022 (Shrider & Creamer, 2023).
International Comparisons of Child Poverty Rates
The income gaps among children living in low-income, middle-class, and affluent families are larger in the United States than in other industrialized Anglophone countries. This is most evident in international comparisons of child poverty provided by the Organization for Economic Cooperation and Development (OECD), which typically use a poverty line defined by a certain fraction—often 50%—of each country’s median income. In the case of the United States, median household income was about $68,000 in 2020. A poverty line based on 50% of median income would thus be $34,000, so any child living in a household with income below that amount would be counted as poor. In contrast to the poverty threshold approach to tracking poverty in the United States, these international comparisons rely on a relative definition of poverty based on the fraction of families whose income is low relative to overall income in the country.2
___________________
2 National Academies (2019a) also provides international data on absolute child poverty, which involve translating the poverty thresholds in the United States, measured in dollars, into comparable thresholds based on other countries’ currencies. Countries with the lowest per capita income will tend to have higher child poverty rates using absolute (rather than relative) poverty measure. As a result, relative to the United States, absolute child poverty is found to be higher in the United Kingdom and lower in other Anglophone countries. U.S. child poverty rates for 2021 are likely to compare more favorably to other countries because of the resources provided to low-income families through the CTC expansion and other pandemic relief measures.
Estimates of child poverty defined in this way for the United States and four English-speaking comparison countries are shown in Figure 6-2. At 20.9%, the rate of child poverty is much higher in the United States than in these peer countries—more than twice as high as in Ireland and more than 5 percentage points higher than in Canada, the country with the second-highest child poverty rate. More detailed analyses trace most of these differences to the much lower fractions of gross domestic product spent on safety-net programs such as child allowances in the United States relative to the comparison countries (National Academies, 2019a, Chapter 4).
Conclusion 6-1: When tax credits, Supplemental Nutrition Assistance Program benefits, and other noncash sources are counted as part of income, the family incomes of children on the bottom rungs of the income distribution have nearly doubled over the past 40 years, and rates of child poverty have been cut in half. The family incomes of children on the middle rungs have grown a bit faster, while those of children on the top rungs have grown much faster. Child poverty in the United States is considerably higher than in other Anglophone countries when poverty is measured by relative income position.
NOTES: Poverty is measured by the Organization for Economic Cooperation and Development-50 poverty rate, which is defined as 50% below each country’s median income. All data are for 2016, the most recent year available.
SOURCE: Data from OECD Stats, https://stats.oecd.org/
Trends in Earnings and Employment
Parents’ employment has always been and remains one of the most important means of avoiding intergenerational poverty. Over the last 50 years, earned income alone lifted the incomes of between 70% and 75%of children above the poverty line (National Academies, 2019a, Figure 4-1). Like family incomes, the earnings of low- and middle-skilled workers have risen more slowly than the earnings of higher-skilled workers (Figure 6-3).3 Between 1973 and 2019, the average hourly earnings for workers at the 10th and 50th percentiles grew by about 25%. Worker wages at the 90th percentile grew by more than twice that amount. The gaps between the earnings of workers in these various groups are generally larger in the United States than in most other industrial countries, and for the most part they have grown more quickly here than elsewhere over the past four decades (Bourguignon, 2022).
Causes of Labor Market Trends
To design effective policies to raise parental earnings among those living below poverty, it is important to understand the causes of low levels of employment and earnings and their impact on inequality. Although scholars continue to debate these causes, they point to three sets of factors in the U.S. labor market:
- Competitive market forces, like technological change and globalization, that have favored higher-skilled workers much more than their lower-skilled counterparts;
- Structural problems in labor markets, such as the fact that firms have more access to labor market information than workers do and the often-high costs to workers of changing jobs, both of which can increase the relative power of employers and lead to lower wages for workers; and
- Weakening laws and institutions, leading to stagnant federal minimum wages and a smaller share of workers who are union members.4
___________________
3 The wage data presented in this chapter end in 2019—just before the onset of the COVID-19 pandemic. The committee judged that it was too soon to interpret wage trends during COVID, given high inflation and large fluctuations in unemployment.
4 These three factors help explain why less of the nation’s productivity growth in recent decades has gone to less-educated U.S. workers than to highly-educated workers and employers, compared with the period right after World War II. Rising health care costs also help account for a growing gap between worker earnings and productivity (Saez & Zucman, 2019); and U.S. productivity growth has also slowed over time, and this slowdown still contributes to the lower earnings growth of all workers (Stansbury & Summers, 2017).
NOTE: Average hourly wages are reported for those who are 16 years or older and are adjusted to $2021 by the PCEPI.
SOURCE: Data from Economic Policy Institute (2022a).
Technological Change and Globalization
Many economists believe that technological changes in the digital era (and earlier) are behind the increasing gap between low- and higher-skilled workers, since many of these changes have favored and rewarded workers with college degrees more than less-skilled workers. Technological changes—and globalization—have restructured the labor market by reducing the amount of “routine” work required by middle- and high-wage jobs. Today jobs in the middle of the wage distribution—in technical aspects of health care or in advanced manufacturing or information technology, for example—usually require some postsecondary skills and credentials, making it much harder for many low-income workers to secure them (Holzer,
2015; Strain, 2020). Increasingly, the labor market also rewards a range of socioemotional skills—such as interpersonal skills and the ability to communicate and work in teams (Orrell, 2021).
Globalization in its many forms—including an expansion of trade in goods and services, as well as immigration—probably reinforces this trend, as good-paying jobs for workers without college degrees in durable manufacturing and other industries have become scarcer as a result of automation and foreign competition (Autor et al., 2008). Labor-market changes favoring higher-skilled workers have laid the groundwork for the impressive successes of a number of career training programs and form the basis for career training policies considered in Chapter 4.
The labor market that children in families living below poverty will enter will continue to evolve, as new forms of robotics and artificial intelligence (AI) become established throughout the economy. Whether the same skills will continue to be compensated well over time remains quite uncertain, as AI gains the ability to perform a much wider range of tasks that have previously required human capabilities (Autor, 2022). Workers will need a solid base of general skills—often referred to as 21st century skills (National Research Council, 2012)—if they are to adapt to changes in the specific skills that the labor market demands and rewards at a given time and place.
Structural Problems in Labor Markets
A second set of explanations for low wages and labor-market inequality focuses on non-competitive forces in the U.S. labor market, which give employers power to set lower wages than they could if the labor market were fully competitive.5 The importance of these factors is reflected in the substantial difference between the earnings of apparently comparable workers who are employed by different firms. Many, but not all, firms have aggressively cut wages, particularly for less-educated workers. At the other end of the spectrum are firms that choose to pay above-market wages and create “good jobs” by investing heavily in worker skills and performance (Osterman, 2018; Ton, 2014). Indeed, high-wage firms can create a “public good” by raising worker productivity and wages. Economists argue that public goods might merit public support (e.g., in this case subsidies or technical assistance)—although there is little evidence to date on the effectiveness of strategies to encourage the creation of “good jobs.”
___________________
5 Economists refer to the power employers have to lower wages when they face limited competition for workers as “monopsony power.” Monopsony can have many causes, ranging from a literal market concentration of employers (e.g., in company towns) to frictions in the job search process or anti-competitive collusion among employers through tools such as non-compete agreements (Card, 2022; Nunn & Hunt, 2021).
Weakening Laws and Institutions
A third set of explanations involves institutional factors, including legislation, that may have slowed wage growth over time, particularly for workers without college degrees. Historically, labor unions have helped to balance power in the labor market and enabled workers to bargain effectively with their employers. Recent studies indicate that unions have increased members’ wages by about 10% to 20% (Farber et al., 2018). However, the prevalence of unions in the United States has declined precipitously. Today only 11.6% of U.S. workers represented by unions (Macpherson & Hirsch, 2021), as compared with 30% of U.S. workers in the 1950s and the current average of 33.1% in countries belonging to the OECD (2019). U.S. unionization is increasingly concentrated in the public sector with only 7.0% of private-sector workers represented by unions. Given the relatively high wage rates of unionized workers, the relevance of unions for the economic status of parents who would otherwise be experiencing poverty remains subject to debate, since it has always been very difficult for unskilled workers in the bottom decile or quintile of earnings in the United States to unionize (Hirsch, 1980).
Legislatively mandated minimum wages have failed to keep pace with the cost of living, contributing to wage stagnation among lower-skilled workers in general and workers of color in particular (Brown & Hamermesh, 2019; Derenoncourt & Montialoux, 2021). The 2022 federal minimum wage was $7.25 per hour, its lowest value in real terms (using the personal consumptions expenditure inflation index) since 2007 (and, before that, since 1996).6 By setting minimum wages above the federal level, a number of states and some cities have offset this decline to some extent, although many U.S. jobs are in states and localities that have not raised the statutory minimum above the federal level.7
Economists continue to debate whether higher minimum wages reduce employment among the youngest and least-educated workers—and if so, by how much (Burkhauser et al., 2023; Cengiz et al., 2019). Some evidence (Clemens & Strain, 2021; Sorkin, 2015) suggests that minimum wage increases have larger long-run negative effects, and that large increases are particularly problematic, although there is disagreement about these conclusions as well. Economists also disagree about the extent to which higher minimum wages can reduce poverty.
___________________
6 Using the Consumer Price Index to measure inflation, the real value of the minimum wage is now at its lowest level since the 1950s.
7 In 2022, more than 30 states raised their minimum wages above the federal statutory level of $7.25, as have several major cities (New York, NY, Los Angeles, CA, Seattle, WA, San Francisco, CA, Washington, DC, and many others).
Conclusion 6-2: Low wages among less-educated workers (including lower-skilled workers who are parents) over the last several decades in the United States can be attributed largely to three factors: competitive market forces, such as technological change and globalization, which have increased skill requirements for middle-class jobs; structural problems, as limited information for employees and the costs associated with changing jobs have strengthened the bargaining power of employers and reduced the power of employees; and weakening laws and institutions, such as federal minimum wages and unionization.
Causes of Unequal Employment and Earnings by Gender and Race
Both wage levels and employment rates show important differences by gender and by race. Figure 6-4 shows the median wages of full-time wage and salary workers by race and gender, while Figure 6-5 presents employment rates for these groups. They show that:
- Median wages among Black and Latino workers are well below the wages of White workers, and this holds true for both men and women. At the same time, wages for women are below men’s wages for all subgroups.
- Employment among women continues to lag behind that of men within each racial group.
- Employment of Black men lags substantially behind that of other men.8
While the wages and employment rates of women relative to men improved greatly during the second half of the 20th century, progress largely slowed or ceased after 2000 (Blau & Kahn, 2017). Many economists believe that the conflicting demands of employment and caregiving responsibilities contribute to this persistent gender gap (Black et al., 2017; Goldin, 2021). This is a particular challenge for low-wage women, since the cost of child care relative to their wages can be very high (Borowsky et al., 2022).
___________________
8 Figure 6-5 shows employment rates for the “noninstitutional” population, which excludes inmates. In 2018, 1.5% of Black men were incarcerated—a rate that was twice as high as for Latino men and five times as high as the White rate. https://www.pewresearch.org/facttank/2020/05/06/share-of-black-white-hispanic-americans-in-prison-2018-vs-2006/. Low-income Black men in the population are also significantly undercounted. As a result, the official employment counts for Black men are biased upwards, and racial employment and labor-force gaps are actually wider than those shown in Figure 6-5 (Holzer, 2021).
NOTES: Data are for wage and salary workers ages 16 and older. Persons whose ethnicity is identified as Hispanic or Latino may be of any race. Data on Native American and Alaska Natives are not reported.
SOURCE: Data from the U.S. Bureau of Labor Statistics (2023).
Policies that increase maternal employment could have deleterious effects on child development if parents are unable to secure high-quality child care. Rigorous evidence shows that expansions of the federal Child Care and Development Fund’s child care subsidies prior to the revision of that program in 2014 increased maternal employment, but also may have worsened child development outcomes (Herbst & Tekin, 2010; see also Baker et al., 2019, and Chapter 4). The 2014 revisions expanded access to higher quality child care, but whether they have improved child outcomes remains uncertain (Johnson & Ryan, 2015). Increasing the generosity of child and dependent care tax credit payments to low-income parents may also increase employment, although the impacts of those types of increases have not been studied.
Substantial disparities in both employment and earnings remain between White workers and both Black and Latino workers, at least partly because of discrimination (Charles & Guryan, 2008; see also Chapter 3).
NOTE: Data show the share of the civilian noninstitutional population ages 16 and older who are employed.
SOURCE: Data from Economic Policy Institute (2022b).
Owing to low employment and labor force participation rates, the annual earnings of Black men are lower than those of White men (Bayer & Charles, 2018; Holzer, 2021)—especially in the case of previously incarcerated people. The lower employment rates and earnings of Black men reentering the labor market after incarceration reflect their lower skills and work experience, but also their lack of access to social networks that might help them get jobs and, as confirmed in a number of rigorous studies, clear discrimination by employers (Holzer et al., 2003; Pager, 2003). Employers tend to fear poor work performance or even criminal conduct, even when the risks of recidivism are very low (Bushway et al., 2022). However, tight labor markets are known to reduce discrimination against people of color (Holzer et al., 2006) and vulnerable workers, including those with criminal records or disabilities (Shambaugh & Strain, 2021).
Conclusion 6-3: Earnings and employment gaps by gender, especially among less-educated women, can probably be attributed in part to difficulties securing affordable and high-quality child care. Gaps by race, especially between Black and other groups of men, reflect differences in skill and experience, but also reflect discrimination and other barriers to employment. Discriminatory barriers appear to be especially severe for previously incarcerated men, particularly Black men.
DO FAMILY INCOME, PARENTAL EMPLOYMENT, AND EARNINGS DURING CHILDHOOD DRIVE INTERGENERATIONAL POVERTY?
Childhood Poverty and Intergenerational Outcomes
A child growing up below the poverty line experiences worse outcomes, on average, than a child from a nonpoor family in virtually every dimension. In its review of the relevant literature, the National Academies’ 2019 report, A Roadmap to Reducing Child Poverty, documented worse outcomes for children in poverty ranging from physical and mental health to completed schooling and from labor market success to risky behaviors and delinquency (National Academies, 2019a). But the largely correlational evidence on which that conclusion is based does not prove that policy-driven increases in the family incomes of children who are in households living below poverty would reduce intergenerational poverty. Income-based childhood poverty is associated with a cluster of other disadvantages for children, including low levels of parental education and living with a single parent (Currie et al., 2013). Are the differences between the life chances of children living in poverty and other children a product of differences in childhood economic resources per se, or do they stem from these other, correlated conditions?
It is easy to think of reasons why income itself matters. Families with higher incomes are better able to meet their children’s basic needs and to invest in goods and services for their families, such as toys and books that provide cognitive stimulation as well as higher-quality nonparental child care and enrichment activities. Moreover, higher incomes may allow parents to reduce or restructure their work hours to spend more time interacting with their children (Becker, 1991). A complementary “stress” perspective focuses on possible links between economic hardship and parental psychological distress. Psychological distress can spill over into couple relationships and can trigger harsher, inconsistent, and more detached parenting (Brody et al., 1994; Conger et al., 1994). Such lower-quality parenting may, in turn, harm children’s cognitive and socioemotional development (Conger et al., 2002; McLoyd, 1990).
Causal Studies
The National Academies (2019a) poverty report moved beyond the correlational literature to focus its attention on studies that attempted to estimate the causal impacts of childhood poverty on children. It considered studies of poverty impacts on both short- and longer-run outcomes, as well
as studies of programs such as Food Stamps (now SNAP) and Medicaid that were introduced decades ago. The report’s overall conclusion was that
(t)he weight of the causal evidence indicates that income poverty itself causes negative child outcomes, especially when it begins in early childhood and/or persists throughout a large share of a child’s life. Many programs that alleviate poverty either directly, by providing income transfers, or indirectly, by providing food, housing, or medical care, have been shown to improve child well-being (National Academies, 2019a, p. 89).
Of particular interest for the current report is the subset of studies that provide causal evidence regarding intergenerational poverty; for example, those that link poverty in childhood to children’s adult economic well-being. One example is the Aizer et al. (2016a) study of the Mother’s Pension Program, which was rolled out during the first three decades of the 20th century. The study found that receipt of benefits in childhood led to an increase in completed schooling and earnings and improvements in adult health. A more recent example is the Hoynes et al. (2016) evaluation of the county-based roll-out of the Food Stamp program in the 1960s and 1970s, which linked the timing of the availability of Food Stamps during childhood to a strong correlate of adult economic status: adult cardiovascular health. It found that the availability of Food Stamps prior to conception or surrounding birth was associated with substantially lower rates of cardiovascular symptoms in adulthood than when Food Stamps first became available later in childhood or in adolescence.
Both the Aizer et al. (2016a) and the Hoynes et al. (2016) studies provide strong evidence that cash or near-cash programs can reduce intergenerational poverty. In both cases, however, the program benefits were first provided five or more decades ago—at a time when the economic conditions faced by low-income families were far worse than they are today. Those conditions set a much lower bar for improvement than was the case in the more recent past or would be today.
An example of more recent research on the intergenerational impacts of safety net policies is Barr et al.’s (2022) study of the availability of child tax credits like the EITC and CTC very early in life. They took advantage of the fact that the families of children born just before January 1 received the tax credits for the baby (which averaged $1,300) for virtually all the child’s first year of life, whereas families of children born just after January 1 were not eligible for the extra year of tax credit until the final (18th) year of the child’s eligibility. In this case, the total amount of tax credits received for the two groups is the same, but the first group received the extra benefits when the children were infants rather than adolescents. Using data on children
born during the 1980s and 1990s, the study found small but statistically significant adult earnings advantages for the early-receipt group. This study reinforces the conclusions of the Hoynes et al. (2016) food stamp study by highlighting the importance of income in the early years of life. Barr et al. (2022) repeated their analyses using data from school records in North Carolina for children born in the 1990s. Here again they found positive impacts for the extra income in the first year of life, specifically on a number of achievement score and behavioral outcomes.
Studies Focused on the Intergenerational Impacts of the EITC
Yet another group of studies has taken advantage of the fact that, beginning in the mid-1990s, many states enacted supplements to federal EITC payments (which are the same regardless of where recipients live). Using this information, Bastian and Michelmore (2018) found causal links between the timing and generosity of the state EITC supplement and children’s completed schooling, employment, and earnings in early adulthood. Specifically, each additional $1,000 of state EITC availability during adolescence was associated with a 4.2% increase in children’s completing college, a 1.0% increase in their employment, and a 2.2% increase in their early-adult earnings. In contrast to the Barr et al. (2022) tax credit study and the Hoynes et al. study of the roll-out of the Food Stamp program, Bastian and Michelmore (2018) found that EITC-induced income increases in adolescence were more potent predictors of children’s adult outcomes than were income increases earlier in childhood.
As described in Appendix C: Chapter 6, other EITC studies based on variation in cross-time or cross-state payment generosity have found impacts on childhood correlates of adult well-being: children’s test scores (Dahl & Lochner, 2012), behavior problems (Hamad & Rehkopf, 2016), college enrollment (Manoli & Turner, 2014), birthweight (Hoynes et al., 2015), and food insecurity (Batra & Hamad, 2021). EITC-based studies generally show both short-run and intergenerational impacts for children, but they differ in their conclusions as to whether effects are larger for EITC payments received while young or later in childhood. Some EITC-based studies suggest that the benefits of the EITC to families and children may be larger for Black individuals than other recipients. For example, Komro et al. (2019) and Batra et al. (2022) found larger effects of state and federal EITC benefits for birth outcomes among Black mothers. Racial/ethnic differences in the literature linking EITC receipt to adult outcomes have not been reported.
When looking at this EITC evidence, it is important to bear in mind that the program creates powerful work incentives for most recipients, and research has shown that it does in fact increase employment among poor single mothers (Nichols & Rothstein, 2016; Schanzenbach & Strain, 2021). The program’s simultaneous impacts on both household income and parental employment make it difficult to determine just how much of the program’s beneficial impact on children comes from its boost to household resources and how much from an increase in parental employment or work hours.
Other Noteworthy Causal Studies of the Intergenerational Impacts of Income Supplementation
Among the relatively few recent studies of the possible intergenerational benefits of increases in family income, in the absence of increases in employment, is the Jacob et al. (2015) comparison of children in families that won the lottery to receive Section 8 housing vouchers in Chicago with children in families that lost that lottery. These researchers found virtually no differences between the two groups in educational achievement and attainment, criminal involvement, or health care utilization, even in the case of children who were very young at the time of the lottery.
Other researchers have taken advantage of actual lotteries that are structured to pay out large prizes in installments over many years, thus creating random variation in household income between lottery winners and losers. Cesarini et al. (2017) find that Swedish lottery wins lead to more health care utilization by winners’ children but have no effect on academic performance. Bulman et al. (2021) examine impacts of lottery wins on college attendance in the United States. Impacts are modest at best and no larger for children in low- than those in high-income families, even when the money was won when the children were young. This suggests no more than a modest role for pure income supplementation, at least with regard to promoting college attendance.
Other studies provide more support for causal impacts of income supplementation. In particular, using data on children from western North Carolina with an oversample of Cherokee Indians, Akee et al. (2010) and Akee et al. (2018) find that unexpected income increases from the distribution of casino winnings to Cherokee youth and their parents reduce symptoms of both behavioral and emotional disorders and also increase the probability that the Cherokee youth would graduate from high school.9 The
___________________
9 The strength of the evidence of income effects on completed schooling is weakened by the fact that the structure of payouts to youth required that they earn high school degrees prior to age 21.
intergenerational evidence on income supplementation does not allow for unambiguous conclusions; EITC-based studies suggest that children benefit when income supplementation is coupled with increases in parental employment, but they may or may not benefit from pure income supplementation.
A different interpretation of the evidence concludes that children may benefit more when parents perceive that the income supplementation is intended for their children. The CTC clearly encourages that perception, and qualitative accounts of parents who receive the EITC also suggest that many believe that EITC dollars are “kids’ money” that should be disproportionately directed toward children’s consumption needs (Sykes et al., 2015). Emerging evidence from a clinical trial of a monthly cash supplement labeled “4MyBaby” indicates that parents allocate a much larger fraction of the cash supplement, relative to other income sources, to child-related expenditures such as books and toys (Gennetian et al., 2022). Messaged connections between income supplementation and child expenditures are weaker in the case of lotteries and assistance from programs providing Housing Choice Vouchers.
Conclusion 6-4: Evidence suggests that income transfer programs during childhood and adolescence have the potential to improve children’s educational and labor market attainment, as well as their physical health, in adulthood. Studies examining policy changes over the past 30 years provide the strongest evidence for intergenerational impacts of expansions of the Earned Income Tax Credit, which increases both employment and income.
Parental Employment and Intergenerational Outcomes
As noted above, programs like the EITC increase both family incomes and parental employment, and it is unclear which effect is responsible for its beneficial impacts on children. Studies of the effect of incomes on intergenerational poverty are reviewed above. Here, we focus on parental employment. Higher levels of parental employment might reduce intergenerational poverty by
- Increasing childhood access to public benefits from the work-based safety net,
- Creating positive parental role models and social networks that boost the future employment and earnings of children, and
- Increasing neighborhood employment or earnings, thereby perhaps expanding children’s access to job networks and role models.
These are reviewed in turn.
Increased Access to Work-Based Safety-Net Benefits
Because so much of the U.S. safety net is tied to work, increasing employment and earnings among parents living below poverty may also increase family income by increasing the family’s access to benefits from programs such as the federal EITC and state supplements to that program. As shown above, strong evidence links parental receipt of the EITC to improved child outcomes and children’s earnings in adulthood, presumably through some combination of increased parental income, employment, and earnings (during the EITC’s phase-in range of income).10 On the other hand, employment-based benefits are, by design, not provided to low-income families whose members are unable or unwilling to work, which can limit the intergenerational mobility of children growing up in these families.
Role Modeling
Parental employment might benefit children by creating positive role models or by expanding parents’ social networks in ways that improve their children’s access to better jobs. On the other hand, children’s development, especially in the case of young children, might be compromised if children are placed in substandard child care or if the often-erratic work schedules of parents provide less continuity at home and limits the time parents can spend interacting with their children. While there is very little evidence on the mechanisms of role modeling and social networks, the possible benefits or harms to children from parental employment per se have been studied for years. On balance, this literature shows neither strong benefits nor harm to children from parental employment.
___________________
10 Barr et al. (2022) show impacts of the EITC on children’s earnings in adulthood that are larger than can be attributed solely to increases in parental employment. On the other hand, Dahl and Lochner (2012) estimate education gains that are larger than can be attributed to earnings and income increases alone. Some policy makers consider the EITC and other work-based public benefits (like subsidized child care) to be preferable to unconditional cash assistance to the poor, which can reduce the work effort of parents. The estimated wage and income elasticities of labor supply in the United States suggest that employment will rise when the wages of the poor are subsidized, but fall if these subsidies are replaced with unconditional cash transfers, including child tax credits. Corinth et al. (2021) argue that the CTC expansion of 2021 would be likely to substantially reduce maternal employment. Using similar methods but different assumptions about the magnitude of wage elasticities, Goldin et al. (2021) estimate a much smaller labor supply response. Ananat (2022) showed very small, if any, negative effects on employment rates associated with the increases in the CTC of 2021, though a permanent increase in the CTC might have larger employment effects than the temporary 2021 credit; and Baker et al. (2021) find virtually no negative effects of child allowances on maternal employment in Canada.
Some of the most rigorous evidence on the effects of parental employment comes from random-assignment evaluations of state welfare reforms made during the 1990s. Some reforms increased maternal employment, but because benefits were reduced dollar-for-dollar as earnings increased, they did not raise family income. Others boosted both maternal employment and family income. There was no evidence from the first set that income-neutral increases in maternal employment either hurt or helped children’s school achievement. Consistent with the EITC-based evidence reviewed above, however, the second set of reforms—which increased both work and family income—produced positive benefits for children, at least in the short run (Duncan et al., 2011). Schildberg-Horisch (2016) provides a more general review of the literature and concludes that most studies fail to find any effect of maternal employment per se—positive or negative—on a child’s short- or long-term educational attainment.
Some evidence suggests that parental employment may have different impacts on child outcomes by child gender and that a mother’s employment very early in the life of a child can have negative impacts on the child (Hill et al., 2005). Furthermore, involuntary employment loss among mothers and fathers may have different impacts on children: Mothers spend more time with their children after losing a job, which offsets the negative effects of their lost income. In contrast, loss of employment does not appear to induce fathers to spend more time with their children or to improve child outcomes (Lindo et al., 2013; Page et al., 2016).
Neighborhood Employment Effects
As discussed in Chapters 2 and 8, increasing parental employment at the neighborhood level might benefit children in the ways suggested by Wilson (1987, 1996)—perhaps through peer effects on behavior as well as through social networks. Correlational evidence shows positive associations between neighborhood employment and the mobility of children later in life (Chetty & Hendren, 2018; Chetty et al., 2016), but there has been no rigorous research on this question.
Conclusion 6-5: Higher parental earnings and employment among low-income families can potentially reduce intergenerational poverty by raising family income, increasing access to the Earned Income Tax Credit (EITC) and other safety-net benefits, and—at both the family and neighborhood levels—providing positive role models and access to good jobs through social networks. Interventions such as the EITC that promote employment and increase income improve children’s long-run outcomes; interventions that promote employment in the absence of
increased income do not appear to improve child outcomes; and evidence on whether income supplementation alone improves long-term child outcomes is inconclusive, with some studies showing positive effects and others showing no improvement.
WEALTH AND INTERGENERATIONAL POVERTY
Wealth and income are distinct from one another, and each may play an independent role in the perpetuation or alleviation of intergenerational poverty. A common economic measure of wealth is net worth, which is the sum of a household’s financial assets (e.g., savings) and nonfinancial assets (e.g., housing equity) minus the sum of all debts or liabilities. In other words, net worth is a point-in-time accounting of assets and liabilities. In contrast, household income refers to the flow of economic resources received by a household over some period, typically a calendar year. Elderly-couple households often have relatively modest annual incomes but may have accumulated substantial net worth, particularly if they have owned a house in an area with rapidly rising housing prices. The families of children, particularly young children with parents in the early stages of adulthood, often have both modest incomes and low (or even negative) net worth, when total debt surpasses total assets, for example if they have accumulated debt to finance their postsecondary education.
Black and Latino families have much less wealth, on average, than White families (Figure 6-6). In the case of Black families, some of these differences may be traced to discriminatory policies like redlining that prevented the accumulation of wealth among Black people (Chapter 3). Black households hold less than 3% of the nation’s wealth, despite constituting 14% of the nation’s population. Their wealth position is also more fragile. Moreover, they are significantly less able to transfer an improved wealth position to their children (Pfeffer & Killewald, 2019). Kerman and Wong (2021) also show that distressed sales of homes, often associated with unstable household income, contribute significantly to the gap in housing wealth between Black and White families.
In the short term, wealth can buffer families against unexpected events such as unemployment, whereas wealth accumulated over the life course can provide economic security in older age or be passed on to children. In the case of intergenerational poverty, low-income families with higher net worth may be able to use their wealth to pay for their children’s education or provide a more stable, safer, or healthier environment for them as they grow up; this, in turn, might reduce the children’s chances of being low
SOURCE: Data from Federal Reserve Board, 2019 Survey of Consumer Finances. https://www.federalreserve.gov/econres/scfindex.htm
income in adulthood. By the same token, wealth-augmenting interventions such as “baby bonds” (described below) might help parents overcome financial obstacles that stand in the way of their children’s success.
Because surveys rarely ask about wealth and its components, wealth is rarely included in discussions of child poverty. Income and wealth are often presumed to be highly correlated, and a low-income household is also assumed to be a wealth-poor household. Using a definition of “net-worth poverty” defined as possessing assets worth just one-fourth of the federal poverty line and data from the Surveys of Consumer Finances, Gibson-Davis et al. (2021) find that three times as many households with children are in poverty when it is defined by net worth rather than income, and that the fraction of households that are in poverty based on both net worth and income is considerably higher among Black (16%) and Latino (14%) households with children than White households (5%).
Household Wealth and Intergenerational Outcomes
Correlational Studies
Parental wealth or its components are highly predictive of child outcomes, based on a number of correlational studies. Most studies focus on postsecondary schooling outcomes, since parents may use their financial resources to help their children attend postsecondary institutions, contribute
to meeting college expenses, and incur debt on behalf of their children (Hotz et al., 2021). This may be a contributing factor explaining why college attainment has also become increasingly unequal, as children from higher-net-worth families are more likely to attend college and graduate with 4-year degrees (Pfeffer, 2018). In an international context, Karagiannaki (2017) in Britain as well as Hällsten and Pfeffer (2017) in Sweden found similar relationships when they examined parental wealth and adult child educational and labor outcomes.
The amount that parents are able to save for their children’s college is positively associated with household income. About one-third of households with less than $35,000 in total annual income reported saving for college; this was true of almost all (91%) households making more than $150,000 (Black & Huelsman, 2012). Parental wealth is associated with increased college enrollment and attendance (Conley, 2001; Doren & Grodsky, 2016; Williams et al., 2009), choosing a 4-year selective college over a 4-year nonselective one (Jez, 2014), and with less undergraduate student debt (Addo et al., 2016; Jackson & Reynolds, 2013). Parental savings earmarked for higher education are associated with increases in attendance (Charles et al., 2007) and college retention (Elliott & Beverly, 2011), as well as reductions in the child’s education debt (Elliott et al., 2014).
Children from wealthier families are more likely to attend college and complete their degrees (Conley, 1999; Jackson & Reynolds, 2013). It is also very common in the United States for parents to provide financial transfers to their children for educational purposes, which are facilitated by greater familial wealth (Taylor & Meschede, 2018). When wealth was disaggregated, liquid assets predicted the likelihood of college graduation (Zhan & Sherradan, 2011), and unsecured debt was negatively associated with graduating (Nam & Huang, 2009; Yeung & Conley, 2008; Zhan & Sherradan, 2011).
Fewer studies focus on links between parental wealth and child labor outcomes in adulthood. Toney and Robertson (2021) connect familial wealth to children’s income returns and document a positive correlation between parental and grandparental wealth and younger generations’ income levels. Fox (2016) found heterogeneous associations between parental wealth and income mobility among Black and White households, with increasing upward income mobility for low-income White families but not low-income Black families.
Causal Studies
Evaluating the likely impacts of wealth interventions requires an understanding of the causal relationship between parental wealth and child outcomes. Causal studies are typically based on the effects of wealth changes that are beyond the control of individual households. One approach is to relate changes in local housing market values to child outcomes. Lovenheim (2011) and Lovenheim and Reynolds (2013) took this approach, focusing on the consequences of housing market–induced changes in parental wealth a few years before children would be expected to start college. Lovenheim (2011) found that students from low- and middle-income families were more likely to attend college when families experienced gains in housing wealth. An additional $10,000 of housing equity increased college attendance by 6 percentage points. It was also associated with families selecting higher-quality schools, and children graduating with a 4-year degree. Lovenheim and Reynolds (2013) found that an increase in home equity is associated with sending children to flagship universities over non-flagship schools. Using a similar causal-modeling approach, Hotz et al. (2021) found that greater housing equity increased college attendance and completion, but did not reduce children’s student debt. Cooper and Luengo-Prado (2015) found that parental wealth increases children’s labor income when they become adults.
Because lotteries can be thought of as a wealth transfer paid in monthly or annual installments, the aforementioned Bulman et al. (2021) study of lottery-winning households is relevant to the discussion of the effects of wealth and income. As discussed above, it found no more than weak linkages to college attendance among children of low-income parents, and it failed to generate strong support for a wealth-effect hypothesis.
While these studies come closer to estimating causal relationships, there are reasons to be concerned about whether the results for lottery winners or homeowners generalize to low-income families. All told, the literature offers no strong evidence as to whether increases in the wealth of low-income families might be expected to reduce intergenerational poverty.
Conclusion 6-6: Controlling for income, family wealth is strongly correlated with children’s adult outcomes. There is mixed evidence on the causal impact of wealth transfers on children’s long-run outcomes. Black families, having significantly less wealth than White families, are more likely to be both income- and wealth-poor, and more likely to experience downward intergenerational wealth mobility. No causal studies have examined differential wealth impacts by race.
INTERVENTIONS INVOLVING CHILDREN’S FAMILY INCOME AND WEALTH AND PARENTAL EMPLOYMENT
The committee’s key task is to use the evidence reviewed above to identify income- and employment-based policies and programs with the potential to reduce intergenerational poverty. As explained in Chapter 1, we characterize the evidence on some of the programs or policies as “strong” and denote them with an “*.” This indicates that the program’s or policy’s impact on intergenerational poverty is supported by random-assignment evaluation evidence that has been replicated across several sites or by compelling quasi-experimental evidence based on national or multi-state data or a scaled-up program. In this case, the evidence supporting all three of our direct-evidence policy ideas was judged to be “strong.”
However, this section also lists policy and program ideas for which supporting evidence is indirect and therefore less certain.
Policy and Program Ideas Based on Direct Evidence
Expanding the Earned Income Tax Credit
The committee’s review of the literature on the intergenerational impacts of policy-induced changes in children’s family income and wealth, as well as parental employment, reveals several promising approaches, but few have been proven to be effective. The strongest direct evidence on the likely intergenerational effects for children is found for programs that increase both family income and parental employment during childhood and adolescence. This leads us to consider potential expansions to tax credits that would produce such increases. The policy research literature does not tell us from which income levels increases are most effective at reducing intergenerational poverty, nor does it identify the most powerful combinations of increased income and employment for enhancing intergenerational mobility. Accordingly, we present three options for reforms with different features. All are presented as modifications to the EITC program, although it would also be possible to link such reforms to other tax credits (such as the CTC). The first option scales up the existing EITC proportionally. The second concentrates the expansion among very low earners. The third adjusts the credit to provide income to nonworkers while still strengthening the incentive to work that is a key feature of the current EITC.
- EITC Option 1: Expand the EITC by increasing payments by 40% across the entire schedule, keeping the current range of the phase-out region. This option is the most straightforward because
- EITC Option 2: Expand the Earned Income Tax Credit by increasing payments along the phase-in and flat portions of the EITC schedule. This less expensive ($8.5 billion annually) option would increase the phase-in rate and thus expand the incentive to work, while increasing resources available to the lowest-income families with earnings. A version of this idea is also included as “EITC #1” in the National Academies (2019a) report. There is a real-world precedent: California supplements the federal EITC with a payment schedule of this type. If this option is implemented, the policy will need to be designed to address low take-up among low-income filers, since those below the IRS filing threshold often forgo filing taxes (Hamad et al., 2022), and the administrative burdens of filing may therefore limit the access of the lowest-income families that need the benefit the most.
- EITC Option 3: Increase generosity throughout the schedule, as in Option 1, but also make available an annual credit of $1,000 per child to families without earnings, phasing this out as earnings approach the first threshold. A key feature of this policy is that it is available to families without work, which are entirely left out of both the EITC as it currently stands and the first two options. These families tend to have the lowest level of resources, and a transfer to such families was a feature of the 2021 CTC expansion. In this option, the maximum credit is increased sufficiently to ensure that the incentive to work is greater than in the status quo. As for Option 2, the committee considered policy design that could facilitate the access of low-income earners and those with zero income who are not required to file taxes. The precise cost of Option 3 is unknown, although it would be considerably more expensive than either Option 1 or Option 2.
it would simply increase the amounts of existing credits by 40%. It is illustrated in the upper-left panel of Figure 6-7 for the example of a single mother with two children and detailed in Appendix C: Chapter 6. This EITC reform was also included as “EITC #2” in the National Academies (2019a) report. That report estimated the annual cost of this EITC expansion to be $20.4 billion and predicted that it would increase employment among low-income adults by 548,000 and reduce short-run poverty among children by 15%.
SOURCE: Based on committee-created simulations under different policy options for structuring an EITC expansion.
Policy and Program Ideas Based on Indirect Evidence
A number of additional policy and program ideas are supported by indirect evidence, though not by direct evidence. They are listed here, with details provided in Appendix C: Chapter 6.
Combining Expansions of the EITC with Other Programs
To increase the stability of monthly incomes for low-income families with children, the EITC expansion could be combined with other programs. As detailed in Appendix C: Chapter 6, three possible options are as follows:
- Combine an expansion of the EITC with expanding coverage of the Child Care and Development Fund block grant to help address the problem of a mismatch between the timing of child care expenses and EITC receipt during a child’s first year of life.
- Combine an expansion of the EITC with expanding coverage of the Child Care and Development Fund block grant to provide more generous and timely reimbursement for the child care expenses of low-income parents.
- Combine an expansion of the EITC with expanding coverage of the Child Care and Development Fund block grant along the lines of what was implemented temporarily for the 2021 tax year.
Steps to Address Low Wages
The education chapter (Chapter 4) spells out several policy ideas for addressing wage stagnation among low-skilled workers by improving the education and labor market skills of parents in poverty. These include broadening supports for low-income students at community colleges and expanding sector-based training. All these policy ideas are supported by direct evidence on intergenerational effectiveness. Other approaches to reduce inequities associated with growing employer market power and weakening worker institutions, all of which rely on indirect evidence, include these:
- Increase the federal minimum wage to $11 per hour or to $15 per hour, in 2023 inflation-adjusted dollars.
- Encourage the expansion of union representation in the labor force.
- Assist and incentivize the creation by private employers of more good jobs.
To improve labor market opportunities for people of color living below poverty (including those who have been formerly incarcerated):
Improve disadvantaged workers’ access to more or better jobs.
Steps to Increase the Labor Force Participation of Young Mothers
Improving the labor market opportunities of low-income mothers could be accomplished with two policy approaches, which were also included above in the EITC section:
- Expand coverage of the Child Development Block Grant.
- Restructure the Child and Dependent Care Tax Credit to provide more generous and timely reimbursement for the child care expenses of low-income parents.
Steps to Increase the Wealth of Low-Income Households
Attempts to increase the wealth of low-income households and improve the short-term outcomes of these families and longer-term outcomes of their children have focused on encouraging the accumulation of savings, for example through child development accounts, and the acquisition of assets, such as by becoming a homeowner (see Appendix C: Chapter 6). To raise wealth among low-income households, especially those of color, the committee views baby bonds as promising:
- Create baby bonds for children born in the United States, with the value of the bonds determined by the family’s income and net worth at the time of the child’s birth and targeting families with incomes below the SPM poverty thresholds and total net worth in the bottom quintile of the U.S. net worth distribution.
