Using data from the Panel Study of Income Dynamics, this study analyzes the effect of exposure to the Earned Income Tax Credit (EITC) in childhood on marriage and childbearing in early adulthood. Results suggest that EITC exposure in childhood leads women to delay marriage and first births in early adulthood (ages 16–25), but has no effect on men. A $1,000 increase in EITC exposure in childhood leads to a 2%–3% decline in a woman's likelihood of having a first birth and a comparable decline in her likelihood of marrying by her early 20s. We find similar reductions in fertility among Black and White women, though marriage declines are concentrated among White women. Results are focused on children growing up in the bottom half of the income distribution and those who spent the majority of childhood residing with a single parent—two groups that are the primary beneficiaries of the EITC. These findings have important implications for the well-being of individuals exposed to the EITC in childhood, as well as their future children. ## Introduction The Earned Income Tax Credit (EITC) is one of the largest cash transfer programs in the United States, redistributing approximately$65 billion to 26 million households every year (Internal Revenue Service 2018). A long line of research illustrates that the EITC substantially affects low-income families by increasing maternal labor supply (see Nichols and Rothstein (2016) for a review), reducing poverty (Hoynes and Patel 2018), and increasing assets (Jones and Michelmore 2018; Michelmore and Lopoo 2021). A small but growing literature examines how the EITC affects the children of beneficiaries, demonstrating that the EITC increases infant birth weight (Hoynes et al. 2015), improves test scores (Dahl and Lochner 2012), and increases educational attainment (Bastian and Michelmore 2018; Manoli and Turner 2018).

Less is known about the long-term effects of the EITC on other aspects of children's transitions to adulthood, such as their childbearing and marriage as young adults. This is surprising, as there has long been interest in understanding whether social welfare policies affect marriage and fertility decisions (for a review, see Lopoo and Raissian 2012, 2014). Much of this literature examines whether social welfare policies, such as the Supplemental Nutrition Assistance Program (SNAP) or Temporary Assistance to Needy Families (TANF), affect contemporaneous marriage and fertility decisions, even when not the explicit goal of the program. Most reviews of this literature show very little evidence that these policies affect family formation (see, e.g., Lopoo and Raissian 2012, 2014; Moffitt 2003). Previous research has also examined whether the EITC affects the marriage and family formation patterns of its adult recipients (e.g., Baughman and Dickert-Conlin 2009; Dickert-Conlin and Houser 2002; Fisher 2013; Herbst 2011; Michelmore 2018), often finding small, insignificant effects.

To date, the literature has largely ignored the intergenerational effects of the EITC on marriage and family formation—that is, how the EITC may affect the marriage and family formation of the children of EITC recipients. Social policies could have a substantially different effect on the marriage and fertility decisions of the children of beneficiaries compared with adult beneficiaries. Exposure to the EITC in childhood potentially provides many years of increased family income, which could influence marriage and family formation in adulthood through a variety of mechanisms. For instance, previous research has linked the EITC with improved student achievement (Dahl and Lochner 2012, 2017), increased educational attainment (Bastian and Michelmore 2018; Manoli and Turner 2018), higher earnings (Bastian and Michelmore 2018), and better health in adulthood (Braga et al. 2020). Increasing human capital raises the opportunity cost of early childbearing (Ellwood and Jencks 2004; Wolfe et al. 2001) by improving the educational outcomes of those exposed to the EITC in childhood; therefore, the EITC may lead to delays in marriage and childbearing.

Using the Panel Study of Income Dynamics (PSID) and a panel of respondents born between 1968 and 1992, we exploit variation in federal and state EITC generosity to examine how policy-induced increases in exposure to the EITC throughout childhood affect subsequent marriage and fertility when children reach early adulthood (ages 16–25). Given the many federal and state policy changes to the EITC over the period, the amount of EITC benefits to which children are exposed from birth to their teenage years varies widely over time, depending on the state of residence, as well as the number of children residing in the household. Our primary interest is in examining whether EITC exposure in childhood delays the timing of that individual's first birth. Because fertility and marriage are intertwined for many individuals, although less so today than in the past (Lundberg et al. 2016), we also investigate the relationship between EITC benefits in childhood and the timing of first marriage.

Results suggest that a $1,000 increase in EITC exposure during one's childhood reduces the likelihood of having a birth by age 21 by 0.8 percentage points, and the likelihood of marrying by age 21 by 0.4 percentage points—both representing a 3% decline in early childbearing and marriage. These effects are somewhat attenuated by age 25, which is indicative of a change in the timing of childbearing and marriage, rather than a change in total fertility and marriage. We find some support that the delays are because of human capital increases, particularly a greater likelihood of completing college on time. We do not find evidence that EITC exposure affects the marriage or fertility for men through age 25. We find similar effects on the fertility of Black and White women, though the effects on marriage are confined to White women. Our results are concentrated among women who grew up in households with an average family income in the bottom half of the income distribution, and among those who spent the majority of their childhood residing with a single parent—two groups that represent the primary beneficiaries of the EITC. As we will show, these results are robust to a number of different analytic decisions: relying on either the federal or state variation in the EITC alone, varying the range of birth cohorts used in the analysis, varying the age interval used to measure EITC exposure in childhood, and a number of different model specifications to control for state and national trends in fertility and marriage over time. These robustness checks provide further confidence that these results represent the causal effect of exposure to the EITC in childhood, rather than merely a spurious correlation. Together, these results have implications for the well-being of children and the intergenerational transmission of poverty. Given the growing socioeconomic gap in marriage and family formation patterns in the United States (McLanahan 2004), these results suggest that policies that increase family income throughout childhood, such as the EITC, could have a substantial impact on economic inequality by delaying fertility among children growing up in low-income households. ## EITC Background The EITC is a refundable tax credit targeting households with earnings below roughly 225% of the federal poverty line, or about$50,000 in 2018. The federal credit is based on earnings, marital status, and the number of children in the household. With the inclusion of state benefits, which typically piggyback off of the federal credit, the EITC benefit was worth as much as $8,524 in 2019 for a family residing in a state with an EITC worth 30% of the federal EITC (such as New York). Together, the federal and state EITCs could be worth more than 50% of a family's annual earnings, received as a lump sum along with the rest of their tax refund. The EITC structure is made up of three regions: a phase-in region, in which benefits increase as earnings increase; a plateau region, in which benefits remain unchanged as earnings fluctuate; and a phase-out region, in which benefits are reduced as earnings increase. The phase-in rate varies depending on the number of children in the household, from 34 cents per dollar for a household with one child to 45 cents per dollar for those with three children. Benefits phase out at a rate of approximately 22 cents on the dollar. Since its inception in 1975, when the federal credit was worth$400 ($1,973 in 2019 dollars) and the phase-in rate was 10%, there have been several federal and state policy changes to the benefit. In 1991, for instance, a larger credit was established for families with two or more children, relative to those with just one. This increased generosity was phased in over time, such that by 1996, the maximum benefit available to households with two children was nearly$1,500 more than that to one-child families ($2,497 in 2019 dollars). In 2009, the federal credit was further expanded for households with at least three children, increasing the phase-in rate to 45% for those families. By 2019, the maximum federal credit available to a family with three or more children was more than$6,500 (in 2019 dollars). These federal policy changes resulted in a change in the maximum credit available by roughly $4,000 between 1990 and 2007, after adjusting for inflation. A summary of the federal expansions is presented in Table 1 in the online appendix. In addition to the federal credit, 28 states and the District of Columbia have also implemented their own EITCs as of 2017.1 State benefits are usually a fixed percentage of the federal benefit, ranging from 3.5% to 40% of the federal benefit.2 Individuals eligible for the federal EITC are also typically eligible for state EITCs, provided they file a state tax return. States began implementing EITCs in the mid-1980s, but the number of states with EITCs increased rapidly after welfare reform in 1996, when states were given more leeway to use welfare block grants for funding state EITCs. While some of our youngest cohorts would have been exposed to state EITCs as children, as we discuss in more detail in the data section, the vast majority of our variation in EITC exposure in childhood is driven by the federal expansions since 1975. ## Why Might EITC Exposure During Childhood Affect Marriage and Family Formation? While a substantial body of research has examined the question of how program generosity—such as the EITC or cash assistance—affects the marriage and family formation patterns of adult recipients (Bitler et al. 2004; Dickert-Conlin and Houser 2002; Fisher 2013; Michelmore 2018), there is less research on how exposure to these programs in childhood affects family formation patterns in adulthood. The theoretical expectations differ substantially between these two groups. Adult recipients, for instance, face direct incentives (or disincentives) to marry or have children in light of program rules, such as an expected gain or loss in benefits associated with such life changes. Children of program recipients, on the other hand, may benefit from exposure to the program itself and to other associated changes that occur in the household, such as increases in income and parental labor supply. Drawing on a human capital framework, we expect that if EITC exposure in childhood improves education outcomes, this should lead to delays in childbearing (and marriage) by increasing the opportunity cost of having a child at a young age. A long line of research has linked increased educational attainment with childbearing. For example, research on the Social Security Student Benefit Program—which provided large subsidies that covered tuition and other costs of attendance for the children of disabled, retired, or deceased parents during the 1970s and early 1980s—showed that it not only increased participation in higher education (Dynarski 2003), but also delayed marriage and childbearing among women who were eligible (Groves and Lopoo 2018). Other research has examined the impact of exogenous policy changes to mandatory schooling requirements on educational attainment and teen fertility. Studies looking at policy changes in school dropout laws found that increasing the age at which children can drop out of school reduces teen fertility (Black et al. 2008; Geruso and Royer 2018) by 5%–30%.3 On the other hand, a study using school start age as an instrument for years of schooling found no reductions in teen fertility (McCrary and Royer 2011). A growing body of work supports the notion that the EITC is linked with increased human capital for children. Hoynes et al. (2015), for instance, tie exposure to the EITC in utero with lower incidence of low birth weight, which is associated with adulthood outcomes (Currie 2011). Dahl and Lochner (2012, 2017) illustrate that exogenous increases in income generated by policy expansions to the EITC improve contemporaneous test scores of children aged 5–15. More recent work also finds evidence that the EITC is linked with higher educational attainment in adulthood (Bastian and Michelmore 2018; Manoli and Turner 2018). Given these findings, we also expect that exposure to the EITC in childhood could lead to delays in fertility in adulthood. In addition to the human capital pathway, EITC exposure during childhood could affect early childbearing through other avenues. A long line of research shows that the EITC increases labor supply among single mothers (Bastian 2020; Eissa and Liebman 1996; Kuka and Shenhav 2020; Meyer and Rosenbaum 2001; Michelmore and Pilkauskas forthcoming; Schanzenbach and Strain 2021; although for an exception, see Kleven 2019), lifts families out of poverty (Hoynes and Patel 2018), and reduces maternal stress (Evans and Garthwaite 2014). An employed parent may serve as a role model for their child, which might make the child less inclined to begin child-bearing at a young age (Chase-Lansdale et al. 2003; Haveman and Wolfe 1994, 1995). In addition, the reduction in maternal stress from the additional resources could improve parent–child relationships, which might lead to childbearing delays (Chase-Lansdale et al. 2003; Mayer 1997). Having an employed parent could also increase access to private health insurance (Baughman 2005; Baughman and Duchovny 2016), which could lower the costs of contraception. On the other hand, because the EITC induces single mothers to work, it may lead to less supervision and, therefore, higher rates of teen fertility (Bastian and Lochner 2020; Haveman and Wolfe 1994, 1995; Hogan and Kitagawa 1985; McLanahan 1988; McLanahan and Sandefur 1994). Finally, EITC generosity—particularly when children reach their teenage years—may increase the likelihood of having children at a young age, because the credit itself reduces the costs of childbearing. Since the generosity of the credit is linked with the number of children in the household, policy expansions that increase the generosity of the EITC may incentivize childbearing among young women. In short, the link between EITC exposure in childhood and marriage and family formation in adulthood is theoretically ambiguous. If the human capital framework dominates, we expect that policy-induced increases in EITC exposure in childhood would lead to delays in marriage and family formation in adulthood. On the other hand, if the income effect of the EITC reducing the costs of early childbearing dominates, we may see an increase in childbearing among those exposed to larger EITC benefits in childhood. ## Data and Methods We use the Panel Study of Income Dynamics for all analyses (PSID 2017). The PSID originated in 1968 with roughly 4,800 families. It collected information on these families and their descendants annually until 1997 and biennially thereafter until 2017—the latest wave of data currently available. Family-level data were collected in each wave from the head of household, defined as the person aged 18 or older with the greatest financial responsibility for the family. Most of the family units were husband–wife pairs in 1968, and the male was typically assumed to be the head of household. In addition to the family-level data collected in each wave, the PSID produces an individual data file, which includes specific information on the individuals in the PSID separate from the family-level data. For example, beginning in 1985, the PSID began collecting a complete fertility and marital history record for each individual in the PSID. We use both the family-level and individual-level data in this analysis. Our analytic sample consists of all individuals born between 1968 and 1992. We restrict our sample to those born before 1993 so we can generate a marital and fertility history of all individuals in the sample until age 25.4 We chose 1968–1992 birth cohorts because we can observe these individuals in the PSID from birth, and because the youngest birth cohorts were exposed to some of the largest expansions to the federal EITC and, thus, provide substantial variation in EITC exposure over the course of childhood. Even though the EITC was not implemented until 1975, we incorporate birth cohorts as early as 1968 because these children could still have received the EITC in later childhood.5 We follow individuals until (1) they turn 25, (2) the 2017 interview, (3) they attrit from the sample, (4) they attrit from the childbirth and adoption history supplement, or (5) they attrit from the marital history file, whichever comes first. Assessing fertility up through age 25 is particularly relevant for this analysis because childless individuals cannot claim the EITC until they turn 25. However, individuals of any age can claim the EITC if they have at least one qualifying child and no one else can claim the tax filer as a dependent on their tax return. Thus, any births averted before age 25 also imply a reduction in the number of individuals eligible for the EITC. We estimate the influence of EITC exposure on several family formation variables measured at different points in the individual's life course, and hence the sample sizes for models estimating the different outcomes vary. The largest sample, however, is 6,622 for models evaluating births and marriage by age 16. Given the different life course trajectories, however, we estimate all models separately for men (n = 3,321) and women (n = 3,301). Following Bastian and Michelmore (2018), we derive a weight for each observation by taking the mean weights from the survey years between birth and reaching age 18.6 All monetary measures are inflated to 2016 dollars using the Consumer Price Index for All Urban Consumers Research Series. ### EITC Measure To analyze how exposure to the EITC in childhood affects childbearing and marriage in adulthood, we calculate the maximum federal and state EITC a child's family could receive given the number of children in the household, the state of residence, and the year. We cumulate the maximum federal and state EITC available from each year from a child's birth to age 15, allowing the state and number of children to vary in each year according to family circumstances, following an approach developed by Bastian and Michelmore (2018). Notably, this measure is independent of a family's own income; changes in the maximum EITC are driven by changes in household size, cross-state moves, and federal and state expansions to the credit over time. Under the assumption that cross-state moves and fertility are not influenced by the EITC, of which there is some evidence (e.g., Baughman and Dickert-Conlin 2009; Kuka and Shenhav 2020), variation in childhood exposure to the EITC is plausibly exogenous with respect to a family's own characteristics.7 Since we focus on a sample of individuals born between 1968 and 1992, we capture variation in the federal EITC stemming from its inception in 1975 and each subsequent policy change that occurred through 2007—the year that our youngest cohort turned 15. We focus on the maximum federal and state EITC available in each year, rather than a family's own EITC, because the latter is endogenous to family characteristics that are also typically associated with fertility and marriage patterns. To be eligible for the EITC in the 2017 tax year, for instance, earnings must have been less than$53,930 for a married family with three children, meaning that EITC-eligible families are negatively selected. Children growing up in lower income households also typically have births at earlier ages than children growing up in more affluent households (Penman-Aguilar et al. 2013), which suggests that a naïve regression of fertility on own EITC benefits received throughout childhood would produce a positive relationship—children exposed to larger EITC benefits would be more likely than those exposed to smaller benefits to have a first child at an early age. In contrast, using the maximum EITC in a given year removes this negative selection into EITC eligibility and allows us to estimate a plausibly causal relationship between EITC exposure and family formation.

Figure 1 illustrates the substantial variation in EITC exposure between birth and age 15 among children born between 1968 and 1992. For the sample as a whole, mean EITC exposure from birth year to age 15 was a little more than $38,000 (in 2016 dollars), with a standard deviation of$21,000. Much of the variation in EITC exposure throughout childhood is driven by differences across birth years, since many of the large expansions to the EITC occurred at the federal level (roughly $37,000 out of the$38,000 in average childhood EITC exposure is accounted for by federal rather than state EITC); however, we also have substantial variation in EITC exposure even within cohorts, since some of the federal expansions produced larger credits for households with multiple children than for those with one child. Variation within birth cohort is primarily driven by differences in the number of children residing in the household each year, as well as by differences across and within states over time.

### Trends in EITC Eligibility in Childhood

Because not everyone in our sample was eligible for the EITC, let alone the maximum EITC, we provide some descriptive statistics regarding EITC eligibility for our sample (Table 1). Since the PSID does not collect information on EITC receipt, we calculate EITC eligibility based on EITC rules given the year, state of residence, number of children residing in the household in each year, marital status of the parents, and income of the parents in the household. On the basis of this calculation, about 40% of the men and women in our sample resided in households that were eligible for the EITC at some point between their birth and age 15.8 Not conditioning on eligibility, the average family was eligible for the EITC for 2.53 years, and the average benefit over those years was $2,574 (federal and state EITC combined, not conditional on eligibility, in 2016 dollars), which implies an annual benefit of$1,017 per year.

We also examine trends in EITC eligibility across several subgroups of interest, and find important heterogeneity in lifetime EITC eligibility and generosity. We note large differences, for example, between White and Black individuals in our sample. While only about one third of White individuals were ever eligible for the EITC between birth and age 15, more than two thirds of Black individuals were eligible. We also note striking differences in eligibility and household benefits according to parental education, the share of childhood spent with married parents, and average income over the course of childhood. Eighty percent of children who spent less than half of their childhood with married parents were ever eligible for the EITC, compared with less than 20% of those who spent their entire childhood living with married parents. Similarly, nearly 90% of individuals in the bottom quartile of average childhood family income were ever eligible for the EITC, compared with 10% of those in the top quartile. In addition, those in the bottom quartile spent 5.39 years eligible for the EITC, for a total household benefit of nearly $9,000, compared with those in the top quartile, who were eligible for less than a year (0.17 year), for a total benefit of about$300.

Patterns by parents' educational attainment are not as striking as those by marital status and income quartile, but we still find a significant gradient: individuals whose head of household did not have a high school diploma were three times as likely as those with a college-educated parent to be eligible for the EITC (60% vs. 20%). It is worth noting, however, that one in five individuals with a college-educated parent, and one in three individuals with a parent with some college experience, was ever eligible for the EITC. More highly educated households may gain eligibility for the EITC during economic recessions or during periods of unemployment (Jones 2017). Additionally, less than half of individuals with a high school–educated parent were eligible for the EITC, further illustrating that parental education is a weaker proxy for EITC eligibility than parental income or marital status.

### Outcomes

We observe several different outcomes in our analyses. To capture the timing of childbearing, we use a dichotomous measure indicating whether the individual had given birth by age 16, age 17, and so on until age 25. Figure 2 shows that the timing of marriage and fertility is quite different for men and women, with men having children at older ages. For example, nearly 10% of women had given birth by age 18, compared with 3% of men who had had a child; by age 25, the corresponding percentages were 40% and 27%, respectively.

We use a similar measure for the timing of first marriage. Consistent with previous demographic research (see, e.g., Auginbaugh et al. 2013), we find differences in the likelihood of marriage for men and women: 3.6% of women had married by age 18, compared with less than 1% of men; by age 25, nearly 40% of women had married, as had less than 30% of men. We also combine the first birth and first marriage measures to generate a measure of births by a particular age prior to marriage and a mutually exclusive indicator for births by a particular age when the respondent had been married at an earlier point.9

To explore some of the mechanisms for the findings, we generate an indicator for completing high school by age 19 and indicators for completing college by age 23 or 25. We also use a continuous measure of the number of years of education completed by age 25. Approximately 63% of respondents had completed high school by age 19, with 67% of women and 59% of men finishing by that age (see Table 3 in the online appendix). We see a large gender disparity in college completion by age 25: 31% of women had finished college by age 25, while 22% of men had done so. These differences are evident in the education completed by age 25 measure as well: the mean value is 13.8 years for women and 13.3 years for men.

### Controls

For race and ethnicity, we generate a set of mutually exclusive variables using reports from the head of household in 1968: White; Black; Hispanic; and a very small set labeled “other,” which includes individuals of Asian descent as well as those who were missing race information. The vast majority of the PSID sample is composed of White (nearly 80%) and Black (around 18%) respondents. To control for the socioeconomic status of the family, we use the number of years of education the head of household had completed in the year the child was born; this variable has a range of 0–17 years, with a mean of 12.64 years. We also control for the proportion of time between the child's birth and age 15 when there were one, two, three, four, and five or more children in the home. Another measure is the proportion of years that the head of household was married between the child's birth and age 15.

In addition to the individual measures, we derive several state-level controls measured in the year the individual was 17, to control for other characteristics of the state that might be correlated with marriage and fertility, as well as with state EITC generosity. We measure the maximum Aid to Families with Dependent Children/Temporary Assistance for Needy Families benefit for a family of three to control for the state generosity in social welfare programs, which is about $592 dollars per month. We use the state unemployment rate to capture the business cycle effects (mean = 5.67%) and state minimum hourly wage (mean =$6.94) as a measure of the value of low-skilled work in the state. These measures were collected from the University of Kentucky's Center for Poverty Research (2021) database.

### Empirical Strategy

We focus on a reduced-form, intent-to-treat analysis for our main specification, and in robustness checks, we illustrate that effects are largest among subgroups most likely to be affected by the EITC. To estimate the effect of EITC exposure on family formation, we estimate models of the following form:
$Yist=β0+β1maxEITCist+β2Xist+β3Zst + 17+αs+γt+δst+εist,$

where $Yist$ is the outcome of interest—whether the individual had a first birth by a given age, and whether the individual ever married by a given age, ranging from age 16 to 25. Each outcome is evaluated for a given individual i, born in state s, in year t. $Xist$ is a vector of demographic controls including race, parental education, average number of children residing in the household throughout childhood, and share of childhood spent with married parents.

$Zst$ is a vector of state contextual variables evaluated in the year the individual turned 17, including the minimum wage, the unemployment rate, and the maximum welfare benefits available for a family of three. Controlling for these state-year contextual factors addresses concerns that state EITC generosity is correlated with other factors at the state level. We also control for state of birth–specific quadratic time trends (δst) to control for other changes in the state over time that may be correlated with EITC exposure and may influence family formation. Additionally, we include state of birth and year of birth fixed effects in all of our models. State of birth fixed effects control for other state-level factors that may be correlated with EITC exposure and family formation, such as political ideology or attitudes toward early childbearing. Year of birth fixed effects control for cohort differences in family formation patterns at the national level. Standard errors are clustered at the state level.

Our coefficient of interest is $β1,$ which represents the change in the outcome of interest among individuals exposed to a $1,000 increase in maximum EITC benefits throughout childhood (ages 0–15; in 2016 dollars). With state of birth fixed effects, year of birth fixed effects, and number of child controls in the model, variation in EITC exposure is driven by variation at each two-way interaction: year of birth by state, state of birth by number of children, and year of birth by number of children. For instance, some of the variation is driven by comparing an individual born in Illinois in 1984 with two siblings with an individual born in Illinois in 1990 with two siblings. Another source of variation is driven by comparing two individuals born in the same year, with the same average number of siblings, but who reside in different states. A third source of variation is driven by comparing individuals born the same year in the same state who have different numbers of siblings. Further variation is generated by changes in household size and state of residence among families between birth and age 15. In our sample of men and women born between 1968 and 1992, there is more than a$100,000 range in maximum potential EITC benefits between birth and age 15, generating substantial variation in exposure to the EITC throughout childhood (see Figure 1).

With all of the controls in the model, we assume that there were no other factors correlated with EITC generosity that may have affected marriage and fertility over this time period. Such factors would have to be correlated with the two- or three-way interaction of these fixed effects to threaten our identification strategy: state by year, number of children by year, number of children by state, or number of children by state and year factors. Given our small sample size, as well as the nature of the variation in the EITC, it is not possible to include all of these two- and three-way fixed-effect interactions in the same model, but we do conduct a number of different model specifications to illustrate the robustness of our main findings. In Table 4 in the online appendix, we illustrate that our results are robust to including number of children by year fixed effects and number of children by state fixed effects, using the maximum number of siblings in the household over childhood fixed effects rather than the average number of siblings, and controlling for number of siblings linear time trends, to address concerns that trends in fertility may have changed differentially by family size over this time period.10

We also conduct a number of subgroup analyses to illustrate that our results are concentrated among groups most likely to be affected by the EITC: those residing primarily in households headed by a single parent and those growing up in the bottom half of the income distribution. As we will illustrate, we find no effect of EITC exposure in childhood among the most advantaged children in our sample: those who grew up in the top half of the income distribution, and those residing with always-married, college-educated parents.

## Results

Figure 3 shows results from our main regression analysis—the relationship between EITC exposure in childhood and fertility and marriage in adulthood for women. Each point is estimated from a separate regression and represents the change in the outcome of interest at each age point following a $1,000 increase in EITC exposure between birth and age 15. All point estimates and standard errors for these models are available in Table 5 in the online appendix. Results indicate a negative relationship between EITC exposure in childhood and early childbearing and marriage, particularly between ages 20–24. Increasing EITC exposure in childhood by$1,000 implies a reduction in the likelihood of giving birth by age 21 of 0.8 percentage points, a 3% reduction in fertility in this sample.11 Point estimates remain negative and significant through age 25, but become slightly smaller in magnitude, which suggests that women exposed to larger EITC benefits in childhood may catch up to their peers who were exposed to smaller benefits.12

We also find a slight negative relationship between EITC exposure in childhood and marriage, particularly between ages 22 and 24. A $1,000 increase in EITC exposure in childhood leads to a 0.5-percentage-point reduction in the likelihood of marrying by age 22, which represents a 3% reduction in marriage by that age. Together, these results indicate that EITC exposure in childhood leads women to postpone marriage and fertility in early adulthood, particularly between ages 20 and 24.13 In Figure 4 in the online appendix, we show results for the subset of our sample we can observe up to age 30, those born between 1968 and 1987. While estimates are noisy, the pattern suggests that the fertility and marriage results are indicative of postponing family formation, rather than forgoing marriage and childbearing altogether. We find little evidence that EITC exposure in childhood affects men's fertility or marriage in early adulthood (Figure 4).14 Point estimates for first births are slightly negative, and become increasingly negative between ages 22 and 25, but never attain statistical significance. We also find very little evidence of a significant relationship between EITC exposure in childhood and marriage in early adulthood, although there is a steady positive trend in the point estimates between ages 22 and 25. The lack of a relationship between EITC generosity and marriage and fertility among men is not entirely surprising, since men tend to underreport births in survey data (Joyner et al. 2012). They also tend to partner with younger women, so the effects for men might be delayed until older ages.15 Given the lack of findings for men, we focus the remaining analyses on women. ### Racial Differences Figure 5 presents results separately for Black and White women and suggests that exposure to the EITC in childhood reduces fertility for both groups between ages 20 and 22, by 0.5–1.0 percentage points.16 After age 22, the effects of the EITC on fertility differ by race. White women exposed to larger EITC benefits in childhood remain significantly less likely to have a birth through age 24, while Black women appear to catch up to their peers who were exposed to smaller EITC benefits by the time they reach their mid-20s. We find very distinctive marriage responses to EITC exposure for Black and White women. For White women, the relationship between EITC generosity and marriage parallels that of the fertility effects: a$1,000 increase in EITC exposure in childhood reduces the likelihood of marrying by age 22 by 0.7 percentage points, which is identical to the point estimate on the likelihood of having a birth by age 22. This pattern persists through age 24, with slight evidence of a reversal of the relationship starting at age 25. We find no clear relationship between the EITC and marriage in early adulthood among Black women, with no point estimates larger than 0.1 percentage points in absolute value, which is consistent with the decoupling of marriage and child-bearing for Black women (Gibson-Davis 2011; Lundberg et al. 2016).

We also examine differences in nonmarital and marital births by race (Figure 6). We find evidence of delays in nonmarital births among Black women, while we find significant effects on marital births for White women. As with our main results by race, we see more evidence that these reductions in nonmarital births for Black women are evidence of postponement, rather than an absolute decline in births, while we see that reductions in marital births for White women persist through age 25. Not surprisingly, since we find no evidence that the EITC affects early marriage for Black women, we likewise find none that the EITC affects marital births among them.

### Parental Education, Income, and Marital Status

To test whether our results are driven by individuals who were more likely to have received the EITC in childhood, we conduct subgroup analyses according to several characteristics that are correlated with EITC eligibility: average family income throughout childhood, education of the head of the household, and the share of childhood spent with married parents.17 For simplicity, we present results for a single outcome: having a birth by age 21.

Results from this exercise—presented in Table 2—confirm that the delays in childbearing seen in our main results are driven by individuals who were most likely to have received the EITC as a child: those with a family income in the bottom quartile of the income distribution and those residing with single parents for the majority of childhood. Among those growing up in families with an average income in the bottom quartile of the income distribution, where nearly 90% of children were eligible for the EITC in at least one year of childhood, a $1,000 increase in EITC exposure in childhood is linked with a decline in the likelihood of having a birth by age 21 by 2.1 percentage points—nearly three times the magnitude of the effect on the main sample.18 We find no significant effects among those with a family income in the top quartile of the income distribution, which is expected since only 10% of children growing up in these households were ever eligible for the EITC. Similarly, we find larger effects of EITC exposure in childhood among those who spent less than half of their childhood residing with married parents: a$1,000 increase in EITC exposure in childhood reduces the likelihood of having a birth by age 21 by 1.2 percentage points. We see a much smaller point estimate among those who spent the majority of their childhood with married parents: a reduction of 0.4 percentage points.19 On the other hand, we find similar effects of EITC exposure in childhood on the likelihood of having a first birth by age 21 regardless of whether the head of household had a high school degree. Recall from Table 1 that we found less of a gradient in EITC eligibility by parental education: one third of children living with a parent with some college were ever eligible for the EITC, compared with about 60% of those living with a parent who did not complete high school. However, when we split the sample according to whether the child spent all of childhood with married, college-educated parents (columns 10 and 11 of Table 2), we find no effect of EITC exposure in childhood on fertility by age 21 among those who spent all of childhood with married, college-educated parents. Effects are concentrated among children who spent at least part of childhood with single, non–college-educated parents. Together, these results imply that EITC exposure in childhood has the largest impact on fertility among children growing up in the most disadvantaged households.

### Sensitivity Analyses: Variation by Birth Cohort, Varying EITC Exposure Window

We conduct a number of sensitivity analyses to test the robustness of our results. First, we limit the sample to those born between 1975 and 1992 (i.e., those born after the EITC was implemented in 1975). We also further limit the sample to those born between 1980 and 1992 to test whether our results are robust to restricting the sample to a more recent birth cohort.

Results of this exercise are presented in Figure 7 and suggest somewhat similar patterns in fertility and marriage postponement across the three different sample restrictions. One notable exception is that for the more recent birth cohorts, we find stronger evidence that women exposed to larger EITC benefits in childhood catch up to their peers who were exposed to smaller benefits in childhood, compared with the pattern for the overall sample. Patterns for marriage, on the other hand, are very consistent across the three samples.

We also test whether results are robust to using different age ranges for the EITC exposure measure (see Figure 7 in the online appendix). Limiting the age range of exposure to a narrower window results in a somewhat larger reduction in fertility by age 21, though estimates are less precise.20 Still, this exercise reveals that results are not sensitive to the age range of EITC exposure.

### Mechanisms: Educational Attainment

What explains the reduction in early childbearing associated with EITC exposure in childhood? Previous research has linked the EITC with reduced infant birth weight (Hoynes et al. 2015), higher test scores in childhood (Dahl and Lochner 2012, 2017), and higher educational attainment and earnings in adulthood (Bastian and Michelmore 2018). These pathways could explain why the EITC also reduces early childbearing among young adult women. Education tends to delay fertility, as women typically avoid having children until after completing school. Conversely, an unexpected birth is also likely to reduce future educational prospects. We test for the human capital hypothesis here, by examining whether EITC exposure from birth to age 15 leads to higher educational attainment in adulthood.

We examine three outcomes related to educational attainment: an indicator for whether the individual completed high school by age 19, an indicator for whether the individual completed college by age 23 or 25, and a continuous term representing the total number of years of schooling completed by age 25. For all women (Table 3), we do not find much evidence that the EITC increases the likelihood of completing high school by age 19, but we do find evidence that the EITC increases the likelihood of completing a college degree by either age 23 or age 25. Since effects are larger when we examine college completion by age 23, this implies that the EITC may help speed up the time to degree, which has been increasing in recent decades (Bound et al. 2010). This is consistent with previous research (Bastian and Michelmore 2018; Manoli and Turner 2018) and implies that a $1,000 increase in EITC exposure in childhood increases the likelihood of completing college by about one percentage point. About 20% of our sample completed a college degree by age 23, so this represents a 5% increase in college attainment. Commensurate with the increase in the likelihood of completing college, we also find evidence that the EITC increases the total number of years of schooling by age 25 by about 0.03 year. These effects are concentrated among White women. We find no significant increase in educational attainment as a function of EITC exposure among Black women, though we find some evidence of an increase in high school degree completion by age 19. This educational attainment pattern may also explain the different fertility patterns we find for Black and White women. Our results suggest that EITC exposure leads Black women to delay their fertility only until about age 23, while the effects for White women seem to persist through the mid-20s. These results are consistent with a human capital theory perspective, which suggests that increases in educational attainment increase the opportunity costs of childbearing, leading to lower fertility rates among young women exposed to the EITC. In our sample, approximately half of women without a college degree had a birth by age 23, compared with 5% of women with a college degree—a difference of 45 percentage points. While this difference is not necessarily causal, if we expect that the women induced to complete a college degree behave like their college graduate peers, a 0.9-percentage-point increase in college completion would predict a 0.4-percentage-point decline in the likelihood of having a first birth by age 23 (the difference in fertility between a college graduate and someone without a college degree, multiplied by our point estimate; 0.45 * 0.009 = 0.004). This implies that about half of our main estimate—that a$1,000 increase in EITC exposure in childhood reduces the likelihood of having a first birth by age 23 by 0.8 percentage points—could be explained by inducing women to complete a college degree. Since we do not find much evidence of a reduction in fertility before age 19, it seems plausible that the fertility reduction is driven by an increase in educational attainment rather than the reverse (that reducing teen fertility improves individuals' prospects for higher education).

## Discussion and Conclusions

The EITC has become one of the largest cash transfer programs in the United States and one of the key social safety net programs for low-income families. While a considerable amount of research has investigated the effects of the EITC on the tax filer and their family, less is known about its long-term effects on the children of EITC beneficiaries. Recent work suggests that the program increases the human capital accumulation of children by increasing educational attainment and earnings in adulthood (Bastian and Michelmore 2018; Manoli and Turner 2018). We hypothesized that the EITC may also reduce the likelihood of marrying and having children in early adulthood. While clearly not the intention of the program, our results suggest that exposure to the EITC in childhood significantly reduces the likelihood of having a birth and marrying among women aged 20–24. We find no effect of EITC exposure in childhood on men's marriage and fertility in early adulthood.

We also corroborate previous research indicating that the EITC increases the likelihood of completing a college degree, providing a plausible mechanism through which the EITC affects marriage and childbearing. These reductions in childbearing likely represent fertility postponement rather than a decline in overall fertility, as we find evidence that women exposed to larger EITC benefits in childhood are no less likely to have had a birth by age 30 than their peers who were exposed to smaller EITC benefits in childhood. Further analysis is needed to determine whether there are quantum effects as well—that is, whether the EITC reduces completed fertility.

This study is not without limitations. First, because of data constraints, we focus on marriage and fertility outcomes in early adulthood. While the data requirements for this project are quite extensive, we are still unable to conclude if the EITC altered the timing of these family outcomes (delaying them) or if there is an overall decline in fertility. Understanding the long-term implications on family formation would offer a more complete understanding of the true impact of the program. Second, our estimates are intent-to-treat estimates, and it is difficult to know the actual treatment-on-the-treated response, which also has important policy implications. Third, our study suggests that education is a potential mechanism that explains the marriage and fertility responses we observe. Future work may consider additional factors, such as employment, which should produce similar effects. It could be the case that the EITC increases both sources of human capital.

These results have implications for the well-being of young women growing up with the EITC, as early childbearing is associated with a host of negative outcomes for both mother and child: for example, decreased educational attainment, poor labor market outcomes, increased likelihood of social welfare program receipt, and poor health (Haveman et al. 1997; Hoffman 2006; Mirowsky 2005). From a budgetary perspective, these results also have implications for federal and state spending on the EITC. Each birth averted before age 25 reduces EITC claims, since childless individuals are not eligible for the credit until turning 25. Together, these results further suggest that the EITC is a cost-effective program, improving the human capital and reducing the incidence of early childbearing among those exposed to the credit in childhood.

## Notes

1

We use data only on federal and state EITC policy changes that occurred between 1975 and 2007, the year that our youngest birth cohort turned 15. Over this period, 20 states and the District of Columbia implemented an EITC.

2

California instituted an EITC in 2017 worth up to 85% of the federal EITC, but it has a different benefit structure than the federal credit and phases out much lower in the income distribution than the federal EITC.

3

Similar reductions in teen and early-20s fertility have been found for other education policy changes, such as lengthening the vocational upper secondary track in Sweden by one year (Grönqvist and Hall 2013).

4

In some analyses, we restrict our sample to those born between 1968 and 1987 and assess first-time childbearing and marriage up through age 30, in order to better estimate whether effects represent a delay or a total reduction in overall fertility and marriage.

5

Results are not sensitive to this restriction. We find similar results when we restrict the sample to a more recent birth cohort (e.g., 1980–1992); this is discussed in more detail in Results.

6

We tried alternative weighting rules, and the results were unchanged. We also ran our preliminary models without weights, and our results were nearly identical.

7

To directly address concerns associated with cross-state moves, in some analyses we restrict the sample to individuals who did not move across states at any point in childhood (about 80% of the sample), and the results are quite similar (see Figure 1 in the online appendix) to the main results reported below. Results are also quite similar if we exclude the state variation from our analyses and rely solely on the federal variation in the EITC (see Figure 2 in the appendix). Additionally, we find no evidence that EITC exposure in childhood affects family characteristics (see Table 2 in the appendix).

8

Take-up of the EITC in recent years is typically over 80% (see Jones 2014).

9

Given this definition, marital births here reflect births that occurred after the first marriage. Some women may have married, subsequently divorced, and then had a child, and would still be considered to have had a “marital birth.”

10

Because of our small sample, we are unable to estimate models with state-by-year fixed effects.

11

To gain some sense of the magnitude of this estimate, we calculate a “back of the envelope” elasticity of around −0.43. The income at the 20th percentile of the income distribution in the United States in 2016 was around $24,000 (www.census.gov/data/tables/time-series/demo/income-poverty/historical-income-households.html). Table 5 in Bastian and Michelmore (2018) provides estimates of the impact of a$1,000 increase in EITC benefits on total family income—the first stage in their instrumental variable estimates—for different age ranges (ages 0–5, 6–12, and 13–18). The estimates for the increased income over these periods range from $10,000 to$12,500 and are statistically significant. The most conservative estimate, $10,000, translates into around a$1,667 annual increase (\$10,000 / 6) or about a 6.9% increase in annual income. For our largest estimate of birth (for 21-year-olds), this suggests an EITC benefit elasticity of a birth of around −0.43 (= −0.03 / 0.069).

12

We also tested models regressing the likelihood of having a birth at a specific age rather than by a specific age (see Figure 3 in the online appendix), and these results, though noisy, are consistent with the EITC reducing fertility in the late teens/early 20s, with no effect, and perhaps a slight positive effect on fertility in the mid-20s.

13

Results are similar if we separately estimate models parsing the EITC variation into its federal and state components. See Figure 2 in the online appendix for results.

14

Point estimates and standard errors are presented in Table 6 in the online appendix.

15

For the subset of men whom we can observe to age 30 (those born 1968–1987), we find a positive trend in marriage as a function of EITC exposure in childhood (see Figure 5 in the online appendix). Point estimates, however, are noisy, and significant only at the 5% significance level at age 27.

16

See Figure 6 in the online appendix for models estimated separately for Black and White men. We find some suggestive evidence that EITC exposure in childhood reduces the fertility of Black men and increases the fertility of White men in the early 20s, which could also explain our null findings for the full sample.

17

Income in childhood is calculated on the basis of the earnings of the head and spouse (if present) and averaged between an individual’s birth and age 15. For individuals lacking family income in every year between birth and age 15, we average family income in the years that the individual was present in the PSID.

18

Using pooled models with interactions between EITC exposure and income quartile, we do find that the point estimate for those in the bottom income quartile is significantly larger than the estimates for those in each of the top three quartiles.

19

These estimates are not statistically significantly different in pooled models with interactions between EITC generosity and whether the majority of childhood was spent with married parents.

20

Some individuals in our main sample have zero exposure to the EITC between birth and age five, since the EITC was not established until 1975. This results in a skewed distribution of EITC exposure from birth to age five, which may account for the much larger effects.

## References

Auginbaugh, A., Robles, O., & Sun, H. (
2013
).
Marriage and divorce: Patterns by gender, race, and educational attainment
(Monthly Labor Review report).
Washington, DC
:
U.S. Bureau of Labor Statistics
. https://doi.org/10.21916/mlr.2013.32
Bastian, J. (
2020
).
The rise of working mothers and the 1975 Earned Income Tax Credit
.
American Economic Journal: Economic Policy
,
12
(
3
),
44
75
.
Bastian, J., & Lochner, L. (
2020
).
The EITC and maternal time use: More time working and less time with kids
(NBER Working Paper 27717).
Cambridge, MA
:
National Bureau of Economic Research
.
Bastian, J., & Michelmore, K. (
2018
).
The long-term impact of the Earned Income Tax Credit on children's education and employment outcomes
.
Journal of Labor Economics
,
36
,
1127
1163
.
Baughman, R., & Dickert-Conlin, S. (
2009
).
The Earned Income Tax Credit and fertility
.
Journal of Population Economics
,
22
,
537
563
.
Baughman, R. A. (
2005
).
Evaluating the impact of the Earned Income Tax Credit on health insurance coverage
.
National Tax Journal
,
58
,
665
684
.
Baughman, R. A., & Duchovny, N. (
2016
).
State Earned Income Tax Credits and the production of child health: Insurance coverage, utilization, and health status
.
National Tax Journal
,
69
,
103
132
.
Bitler, M. P., Gelbach, J. B., Hoynes, H. W., & Zavodny, M. (
2004
).
The impact of welfare reform on marriage and divorce
.
Demography
,
41
,
213
236
.
Black, S. E., Devereux, P. J., & Salvanes, K. G. (
2008
).
Staying in the classroom and out of the maternity ward? The effect of compulsory schooling laws on teenage births
.
Economic Journal
,
118
,
1025
1054
.
Bound, J., Lovenheim, M. F., & Turner, S. (
2010
).
Why have college completion rates declined? An analysis of changing student preparation and collegiate resources
.
American Economic Journal: Applied Economics
,
2
(
3
),
129
157
.
Braga, B., Blavin, F., & Gangopadhyay, A. (
2020
).
The long-term effects of childhood exposure to the Earned Income Tax Credit on health outcomes
.
Journal of Public Economics
,
190
,
104249
. https://doi.org/10.1016/j.jpubeco.2020.104249
Chase-Lansdale, P. L., Moffitt, R. A., Lohman, B. J., Cherlin, A. J., Levine, R., Coley, L. D., . . . Votruba-Drzal, E. (
2003
).
Mothers' transitions from welfare to work and the well-being of preschoolers and adolescents
.
Science
,
299
,
1548
1552
.
Currie, J. (
2011
).
Inequality at birth: Some causes and consequences
.
American Economic Review: Papers & Proceedings
,
101
,
1
22
.
Dahl, G. B., & Lochner, L. (
2012
).
The impact of family income on child achievement: Evidence from the Earned Income Tax Credit
.
American Economic Review
,
102
,
1927
1956
.
Dahl, G. B., & Lochner, L. (
2017
).
The impact of family income on child achievement: Evidence from the Earned Income Tax Credit: Reply
.
American Economic Review
,
107
,
629
631
.
Dickert-Conlin, S., & Houser, S. (
2002
).
EITC and marriage
.
National Tax Journal
,
55
,
25
40
.
Dynarski, S. M. (
2003
).
Does aid matter? Measuring the effect of student aid on college attendance and completion
.
American Economic Review
,
93
,
279
288
.
Eissa, N., & Liebman, J. B. (
1996
).
Labor supply response to the Earned Income Tax Credit
.
Quarterly Journal of Economics
,
111
,
605
637
.
Ellwood, D. T., & Jencks, C. (
2004
).
The spread of single-parent families in the United States since 1960
. In Moynihan, D. P., Smeeding, T. M., & Rainwater, L. (Eds.),
The future of the family
(pp.
25
65
).
New York, NY
:
Russell Sage Foundation
.
Evans, W. N., & Garthwaite, C. L. (
2014
).
Giving mom a break: The impact of higher EITC payments on maternal health
.
American Economic Journal: Economic Policy
,
6
(
2
),
258
290
.
Fisher, H. (
2013
).
The effect of marriage tax penalties and subsidies on marital status
.
Fiscal Studies
,
34
,
437
465
.
Geruso, M., & Royer, H. (
2018
).
The impact of education on family formation: Quasi-experimental evidence from the UK
(NBER Working Paper 24332).
Cambridge, MA
:
National Bureau of Economic Research
.
Gibson-Davis, C. (
2011
).
Mothers but not wives: The increasing lag between nonmarital births and marriage
.
Journal of Marriage and Family
,
73
,
264
278
.
Grönqvist, H., & Hall, C. (
2013
).
Education policy and early fertility: Lessons from an expansion of upper secondary schooling
.
Economics of Education Review
,
37
,
13
33
.
Groves, L. H., & Lopoo, L. M. (
2018
).
Federal financial aid and family formation: Examining the Social Security student benefit program
.
Journal of Family and Economic Issues
,
39
,
436
444
.
Haveman, R., & Wolfe, B. (
1994
).
Succeeding generations: On the effects of investments in children
.
New York, NY
:
Russell Sage Foundation
.
Haveman, R., & Wolfe, B. (
1995
).
The determinants of children's attainments: A review of methods and findings
.
Journal of Economic Literature
,
34
,
1829
1878
.
Haveman, R., Wolfe, B., & Peterson, E. (
1997
).
Consequences of teen childbearing for the life chances of children
,
1968
88
. In Maynard, R. A. (Ed.),
Kids having kids: Economic costs and social consequences of teen pregnancy
(pp.
324
341
).
Washington, DC
:
Urban Institute Press
.
Herbst, C. M. (
2011
).
The impact of the Earned Income Tax Credit on marriage and divorce: Evidence from flow data
.
Population Research and Policy Review
,
30
,
101
128
.
Hoffman, S. D. (
2006
).
By the numbers: The public costs of teen childbearing
(Report).
Washington, DC
:
National Campaign to Prevent Teen Pregnancy
.
Hogan, D. P., & Kitagawa, E. M. (
1985
).
The impact of social status, family structure, and neighborhood on the fertility of Black adolescents
.
American Journal of Sociology
,
90
,
825
855
.
Hoynes, H., Miller, D., & Simon, D. (
2015
).
Income, the Earned Income Tax Credit, and infant health
.
American Economic Journal: Economic Policy
,
7
(
1
),
172
211
.
Hoynes, H. W., & Patel, A. J. (
2018
).
Effective policy for reducing poverty and inequality? The Earned Income Tax Credit and the distribution of income
.
Journal of Human Resources
,
53
,
859
890
.
Internal Revenue Service
. (
2018
).
Individual income tax return line item estimates, 2018
Jones, L. E., & Michelmore, K. (
2018
).
The impact of the Earned Income Tax Credit on household finances
.
Journal of Policy Analysis and Management
,
37
,
521
545
.
Jones, M. R. (
2014
).
Changes in EITC eligibility and participation, 2005–2009
(CARRA Working Paper No. 2014-04).
Washington, DC
:
U.S. Census Bureau, Center for Administrative Records Research and Applications
.
Jones, M. R. (
2017
).
The EITC over the great recession: Who benefited?
National Tax Journal
,
70
,
709
736
.
Joyner, K., Peters, H. E., Hynes, K., Sikora, A., Taber, J. R., & Rendall, M. S. (
2012
).
The quality of male fertility data in major U.S. surveys
.
Demography
,
49
,
101
124
.
Kleven, H. (
2019
).
The EITC and the extensive margin: A reappraisal
(NBER Working Paper 26405).
Cambridge, MA
:
National Bureau of Economic Research
.
Kuka, E., & Shenhav, N. (
2020
).
Long-run effects of incentivizing work after childbirth
(NBER Working Paper 27444).
Cambridge, MA
:
National Bureau of Economic Research
Lopoo, L. M., & Raissian, K. M. (
2012
).
Natalist policies in the United States
.
Journal of Policy Analysis and Management
,
31
,
905
946
.
Lopoo, L. M., & Raissian, K. M. (
2014
).
U.S. social policy and family complexity
.
Annals of the American Academy of Political and Social Science
,
654
,
213
230
.
Lundberg, S., Pollak, R. A., & Stearns, J. (
2016
).
Family inequality: Diverging patterns of marriage, cohabitation, and childbearing
.
Journal of Economic Perspectives
,
30
(
2
),
79
101
.
Manoli, D., & Turner, N. (
2018
).
Cash-on-hand and college enrollment: Evidence from population tax data and the Earned Income Tax Credit
.
American Economic Journal: Economic Policy
,
10
(
2
),
242
271
.
Mayer, S. E. (
1997
).
What money can't buy: Family income and children's life chances
.
Cambridge, MA
:
Harvard University Press
.
McCrary, J., & Royer, H. (
2011
).
The effect of female education on fertility and infant health: Evidence from school entry policies using exact date of birth
.
American Economic Review
,
101
,
158
195
.
McLanahan, S. (
2004
).
Diverging destinies: How children are faring during the second demographic transition
.
Demography
,
41
,
607
627
.
McLanahan, S., & Sandefur, G. (
1994
).
Growing up with a single parent: What helps, what hurts
.
Cambridge, MA
:
Harvard University Press
.
McLanahan, S. S. (
1988
).
Family structure and dependency: Early transitions to female household headship
.
Demography
,
25
,
1
16
.
Meyer, B. D., & Rosenbaum, D. T. (
2001
).
Welfare, the Earned Income Tax Credit, and the labor supply of single mothers
.
Quarterly Journal of Economics
,
116
,
1063
1114
.
Michelmore, K. (
2018
).
The Earned Income Tax Credit and union formation: The impact of expected spouse earnings
.
Review of Economics of the Household
,
16
,
377
406
.
Michelmore, K., & Lopoo, L. M. (
2021
).
Exposure to the Earned Income Tax Credit in early childhood and family wealth
.
Russell Sage Foundation Journal of the Social Sciences
,
7
(
3
),
196
215
.
Michelmore, K., & Pilkauskas, N. (Forthcoming).
Tots and teens: How does child's age influence maternal labor supply and child care response to the Earned Income Tax Credit?
Journal of Labor Economics
.
Mirowsky, J. (
2005
).
Age at first birth, health and mortality
.
Journal of Health and Social Behavior
,
46
,
32
50
.
Moffitt, R. A. (
2003
).
The Temporary Assistance for Needy Families program
. In Moffitt, R. A. (Ed.),
Means-tested transfer program in the United States
(pp.
291
363
).
Chicago, IL
:
University of ChicagoPress
.
Nichols, A., & Rothstein, J. (
2016
).
The Earned Income Tax Credit
. In Moffitt, R. A. (Ed.),
Economics of means-tested transfer programs in the United States
(Vol.
1
, pp.
137
218
).
Chicago, IL
:
University of Chicago Press
.
Panel Study of Income Dynamics (PSID)
. (
2017
).
PSID main interview user manual: Release 2017
.
Ann Arbor
:
Institute for Social Research, University of Michigan
.
Penman-Aguilar, A., Carter, M., Snead, M. C., & Kourtis, A. P. (
2013
).
Socioeconomic disadvantage as a social determinant of teen childbearing in the U.S.
Public Health Reports, 128
(
Suppl. 1
),
5
22
. https://doi.org/10.1177/00333549131282S102
Schanzenbach, D. W., & Strain, M. R. (
2021
).
Employment effects of the Earned Income Tax Credit: Taking the long view
. In Moffitt, R. A. (Ed.),
Tax policy and the economy
(Vol.
35
, pp.
87
129
).
Chicago, IL
:
University of Chicago Press
.
University of Kentucky Center for Poverty Research
. (
2021
).
UKCPR national welfare data, 1980–2019
[Data set].
Lexington
:
University of Kentucky
. http://ukcpr.org/resources/national-welfare-data
Wolfe, B., Wilson, K., & Haveman, R. (
2001
).
The role of economic incentives in teenage nonmarital childbearing
.
Journal of Public Economics
,
81
,
473
511
.
This is an open access article distributed under the terms of a Creative Commons license (CC BY-NC-ND 4.0).