Adult Children’s Marital Timing and Upward Intergenerational Financial Transfers: Variation across Gender, Race, and Ethnicity

Abstract

Marriage is a key adult milestone yet is also considered a greedy institution that monopolizes resources, jeopardizing adult children’s financial transfers to aging parents. As the age at which marriage shifts later into adulthood, this study examines whether marriage still operates as a greedy institution, considering the role of marital timing on upward financial transfers to aging parents. Using data from Waves I–V of Add Health and Wave I of the Add Health Parent Questionnaire from 1994 to 2018, we examined whether timing of marriage (never, early, on time, and late) was associated with adult children’s upward financial transfers to their parents, accounting for variation across gender, race, and ethnicity. Adult children who never married were the largest group who gave upward financial transfers to their aging parents (33%). Results suggested that marriage or timing of marriage was a greedy institution among men, women, white, Hispanic, and Asian adults, but not Black adults, suggesting heterogeneity in the role of marriage on family support.

1. Introduction

Parents and children have always provided critical support to one another over the life course. Support exchanged earlier in the life course often benefits children, with this pattern extending well into adulthood. However, adult children become a key source of support to parents as they age (Kalmijn 2019). The reliance of aging parents on their adult children for support may increase in the coming decades, as longer life expectancies in the United States extend the period of overlap between adult children’s and aging parents’ lives (Bengtson 2001). Moreover, adult children’s support of their aging parents may be integral to meeting parents’ needs as parents live longer but not necessarily healthier lives (Boersma 2020; Jindai 2016). While parents have enjoyed gains in life expectancy over the last several decades, children have encountered an extended entry into adulthood, often taking more years to achieve key adult milestones—including marriage—relative to prior cohorts (Settersten and Ray 2010; Vespa 2017). Changing life circumstances, such as delays in entry into adult roles like marriage, may alter aspects of parent–child ties, including the transfer of financial support resources across generations. This study will investigate how an adult child’s marital timing (e.g., early, on time, late, or never) shapes financial support offered to aging parents (e.g., upward transfers). We focus on marriage because marriage is considered a key marker of adulthood that is representative of a child’s emotional and financial stability (Smock and Schwartz 2020).

To investigate how adult children’s marital timing shapes upward financial transfers, this study draws on five waves of data from the National Study of Adolescent to Adult Health (Add Health). In this study, we ask two key questions. First, how does adult children’s marital timing (e.g., early, on time, late, or never) shape their upward financial transfers to their aging parents? Second, does the relationship between marital timing and upward financial transfers vary by the adult child’s gender or racial and ethnic background? This study contributes to prior work in several ways. First, this study addresses how family processes, like intergenerational financial transfers, are shaped by changing circumstances in adult children’s lives by specifically focusing on the child’s marital timing. Second, this study considers the role of marital timing on intergenerational financial transfers, shedding light on life course principles of linked lives and timing (Elder 1985, 1998). Third, this study contributes to current work on the transition to adult roles and intergenerational ties by investigating sociodemographic differences by accounting for the gender as well as racial and ethnic background of adult children, including Black, Hispanic, Asian, and white adults.

2. Background

Parents and children are crucial sources of support to one another over the life course (Fingerman et al. 2012; Rossi and Rossi 1990). Throughout childhood and well into adulthood, parents provide crucial downward transfers of support through time, money, or advice to their children (Kalmijn 2019; Swartz 2009). As parents grow older, children often begin to direct more support toward their aging parents (Kalmijn 2019). Financial assistance is a key type of support exchanged between parents and their children (Kalmijn 2019; Lee and Manzoni 2023; Silverstein 2006). Although children tend to be the most frequent recipients of financial support over the life course, as parents age, children often step in to provide financial resources to help with parent’s expenses, including housing, living, and medical expenses (Kalmijn 2019). Even though children are the steadier recipients of financial transfers across the life course, it is important to unpack the circumstances that shape upward financial transfers from children to their aging parents. This is especially true as life expectancy has lengthened in the United States, increasing years of overlap between parents and their children (Bengtson 2001). Although aging parents are living longer, this does not necessarily mean these years are spent in good health (Boersma 2020; Jindai 2016). Because of this, intergenerational relationships are increasingly salient for ensuring older adults have the support they need in later life. At the same time, many children have experienced a winding pathway into adulthood, often taking longer to achieve roles considered emblematic of adulthood like finishing a college degree, securing full-time employment, entering marriage, or having a family (Settersten and Ray 2010). Delays on the pathways into adulthood may create challenges for adult children who wish to provide financial support to their aging parents.

Financial support exchanged between parents and their children often fluctuates, depending on the needs and life circumstances of each member of the parent–child dyad. Therefore, it is useful to employ a life course framework to better understand how upward transfers may vary over time, specifically drawing upon the life course principles of linked lives and timing (Elder 1985, 1998). Linked lives describe the enduring bond shared by parents and their children over the life course. Linked lives also explain how family relationships and intergenerational transfers of support should be viewed as a heterogeneous process, influenced by the circumstances of each member of the family dyad, in this case focusing on adult children and their aging parents (Bengtson and Allen 1993; Elder 1985, 1998; Lee and Manzoni 2023). For instance, as parents grow older, they may experience health-related changes associated with aging that could necessitate more support from their adult children. Previous research suggests that financial support transferred between parents and children is often based on the need of the recipient and previous transfers of financial support, such that those with greater needs and a history of providing exchanges will receive financial transfers (Lanuza 2020; Silverstein et al. 2002). At the same time, children may acquire new adult roles, like entering a marriage, which may shape their ability to provide financial support to their aging parents.

In this study, to explore the idea of linked lives, we investigate how an adult child’s entry into marriage may shape the child’s ability to provide financial support to parents.

Several prior studies have considered how entering a marriage may shape adult children’s relationships with their aging parents. Sarkisian and Gerstel (2008) posited that marriage functions as a “greedy institution” or that entering a marriage monopolizes a child’s resources, thereby weakening intergenerational ties. Indeed, Sarkisian and Gerstel (2008) found support for the idea that marriage operates as a greedy institution such that married children are less likely to engage in upward or downward transfers of emotional, financial, or practical support with their aging parents. Echoing in another study, Gerstel and Sarkisian (2007) revealed that married adults are less involved in giving and receiving intergenerational care relative to unmarried adults. However, it remains unclear whether marriage continues to resemble a greedy institution in light of historic delays in marriage in the United States since the 2000s (Julian 2022; U.S. Census Bureau 2018).

Today, marriage remains a key marker of adulthood, yet it has become a more selective institution in the United States. Many adults wait to marry until they have cleared an “economic bar”, which can include completing a college degree, securing full-time employment, living independently from parents, and having a steady income (Cherlin 2020; Gibson-Davis et al. 2018; Settersten and Ray 2010; Zhang and Sassler 2022). As such, marriage is considered a capstone that is often representative of stability (Smock and Schwartz 2020). As marriage has shifted toward being viewed as a “capstone” of adult achievements, it is possible that the relationship between an adult child’s entry into marriage and how it reshapes their ties with their aging parent has changed. Marriage is not only symbolic of a child’s entry into adulthood but is also indicative of the child’s acquisition of resources they deem necessary to enter a marriage (Cherlin 2020). Thus, as marriage becomes increasingly delayed in the United States, it is possible that married adult children may have more financial resources to share with their aging parents, as the accumulation of financial resources is often a precursor for entry into marriage. Indeed, a recent study finds a shift in the relationship between a child’s entry into marriage and intergenerational transfers. Min et al. (2021) found that children’s instrumental support (e.g., time) directed toward parents declined after marrying, yet there was no reduction in upward financial transfers among married adult children.

Another key aspect of this study is the timing of the adult child’s entry into marriage. Timing emphasizes expectations regarding when children should enter adult roles, like marriage (Arocho and Dush 2020; Elder 1985, 1998) and how the timing of role transitions can shape parent–child ties (Alwin 2012; Zhang et al. 2024). Age at first marriage in the United States has become increasingly delayed (Julian 2022; U.S. Census Bureau 2018) as many young adults perceive an economic bar that must be met prior to marrying (Cherlin 2020). The wait to clear this “economic bar” prior to marriage is reflected in ages at first marriage in the United States, which have steadily increased over the last several decades. Among men, the median age at first marriage had risen from 28.2 years in 2010 to 30.1 years in 2021. For women, the median age at first marriage had risen from 26.1 years in 2010 to 28.6 years of age by 2021 (Julian 2022; U.S. Census Bureau 2018). Delaying marriage later into adulthood could similarly delay the competing demands that accompany marriage and allow adult children to accumulate more resources to support themselves and their aging parents.

We also investigate whether an adult child remaining never married until their late thirties shapes their ability to provide upward financial transfers relative to one’s married peers. Due to preferences to remain single or cohabit, partner availability, and racial and ethnic discrimination in the dating market (Marsh 2023; Zhang and Sassler 2022), a growing proportion of Americans never marry—as of 2021, 25% of Americans were never married by age 40, the largest proportion since the 1900s (Fry 2023). By race and ethnicity, 46% of Black adults, 27% of Hispanic adults, 17% of Asian adults, and 20% of white adults had never married by age 40 (Fry 2023).

To summarize, young adults who enter marriage later may have greater economic resources and thus will be better prepared to offer financial support to their parents as they age, relative to their peers who marry on time or early. In this way, marriage could become less of a greedy institution when, at first, marriage is pushed later into adulthood. Thus, we expect that children who marry late (e.g., ages 28 and older) will be more likely to transfer financial resources to their aging parents relative to those who enter marriage on time (e.g., ages from 25 to 27) or early (e.g., ages 24 and younger). Following this reasoning, we also predict that children who enter marriage on time will be more likely to offer financial support to their aging parents relative to children who marry early. Although marrying later may increase the availability of resources that adult children can share with their aging parents, married adult children will still face more resource competition than their unmarried peers. Thus, we expect that unmarried adult children will offer more financial support to their aging parents than children who marry early, on time, or late.

3. Variation by Gender, Race, and Ethnicity

The manners in which the lives of adult children are shaped by their entry and timing into marriage may vary based on sociodemographic characteristics, including the adult child’s gender as well as their racial and ethnic background (Keister 2012; Raley et al. 2015).

3.1. Variation by Gender

Women tend to marry at earlier ages relative to men (Zhang and Sassler 2022). Generally, women also provide more caregiving support to their aging parents than men (Silverstein 2006), making it more consequential if the flow of support from the adult daughter to the aging parent is disrupted by circumstances in the adult daughter’s life, like marriage. Although Sarkisian and Gerstel (2008) did not find gender differences when considering whether marriage is a greedy institution, some prior work suggests that marriage monopolizes more of women’s resources relative to men’s (Laditka and Laditka 2001; Rossi and Rossi 1990; Spitze and Logan 1990).

It is unclear whether sons provide greater upward financial transfers to their parents relative to daughters. Due to the persistent gender–wage gap in the United States, in which men tend to earn more than women (Gharehgozli and Atal 2020), it is possible that men substitute parent caregiving with financial transfers rather than time transfers, thus providing greater upward financial transfers relative to women (Finley 1989). However, prior research has found no significant gender differences in levels of financial transfers provided to parents but confirms significant differences based on marital status, with single adults more likely to provide upward financial transfers than partnered adults (Chesley and Poppie 2009). Despite conflicting findings in prior research, we expect that marriage will operate as a greedier institution for women relative to men, regardless of marital timing. Therefore, we expect that women who marry early, on time, and late will be significantly less likely than men who marry early, on time, and late to provide upward financial transfers to their parents.

3.2. Variation by Race and Ethnicity

The race and ethnicity of the adult child and their aging parent may also shape upward financial transfers following marriage. Due to structural and institutional racism, families of color may have fewer financial resources than white families, necessitating greater financial transfers between parents and adult children (Lanuza 2020). Though Sarkisian and Gerstel (2008) found support for marriage as a greedy institution, they did not find that structural factors (e.g., employment status, income, and educational attainment) explained their findings. Instead, the authors suggest cultural factors may be at play, including the racial and ethnic background of respondents. In fact, Sarkisian and Gerstel found higher odds of offering support as well as shared contact amongst African Americans in their sample relative to white Americans (Sarkisian and Gerstel 2008).

Indeed, cultural values such as familism and filial piety play a crucial role in shaping intergenerational ties (Silverstein et al. 2006). Both emphasize prioritization of the family over oneself, including the normative expectation to provide support to one’s aging parents (Silverstein et al. 2006). Familistic values have been identified among Black and Hispanic Americans, while filial piety often characterizes family ties among Asian Americans (Schwartz 2007), with familism and filial piety representing similar value orientations (Schwartz et al. 2010). As such, intergenerational relationships tend to be closer and exhibit greater solidarity in families of color, leading to more frequent upward financial transfers to parents, particularly among Black, Hispanic, and Asian adults (Burr and Mutchler 1999; Lanuza 2020; Lee and Manzoni 2023; O’Brien 2012; Swartz 2009; Zeiders et al. 2016; Zhang and Grant 2023). Indeed, Black, Hispanic, and Asian families are more likely than white families to report a greater sense of obligation to give back to parents as they age, whether through caregiving or financial transfers (Losada et al. 2006; Stack 1983; Yoo and Kim 2014). Nevertheless, these stronger intergenerational ties in families of color may not always translate into upward financial transfers. In particular, prior research suggests that Black and Hispanic families may face structural constraints related to transfers that curtail their ability to offer financial support but compensate by offering instrumental and emotional forms of support (Berry 2006). Despite possible structural constraints, we expect Black, Hispanic, and Asian adult children to offer more upward transfers to their aging parents relative to white adult children regardless of marital timing due to stronger intergenerational ties in families of color relative to white families.

4. Hypotheses

To summarize, our research is guided by the following hypotheses:

Hypothesis 1.

We hypothesize that upward financial transfers will vary depending on the adult child’s timing of entry into marriage among all adult children and across gender as well as race and ethnicity. From this, we have three specific hypotheses.

Hypothesis 1a.

We hypothesize that unmarried adult children will offer greater financial transfers to their aging parents than children who marry early (e.g., ages 24 and younger), on time (e.g., ages from 25 to 27), or late (e.g., ages 28 and older).

Hypothesis 1b.

We hypothesize that children who enter marriage on time (e.g., ages from 25 to 27) will be more likely to offer financial transfers to their aging parents relative to children who marry early (e.g., ages 24 and younger).

Hypothesis 1c.

We hypothesize that adult children who marry late (e.g., ages 28 and older) will be more likely to transfer financial resources to their aging parents relative to those who enter marriage on time (e.g., ages from 25 to 27) or early (e.g., ages 24 and younger).

Hypothesis 2.

We hypothesize that marriage will operate as a greedier institution for women relative to men, regardless of marital timing.

Hypothesis 3.

We hypothesize that Black, Hispanic, and Asian adult children will offer greater upward financial transfers to their aging parents relative to white adult children regardless of marital timing.

5. Data, Measures, and Methods

5.1. Data

To answer our research questions, we use data from Waves I, II, III, IV, and V of the National Longitudinal Study of Adolescent to Adult Health (Add Health) and Wave I of the Add Health Parent Questionnaire. Add Health is a nationally representative survey in the United States that followed 20,745 adolescents who were in the 7th–12th grades at Wave I from 1994 to 1995. Subsequently, Wave II was collected from 1995 to 1996, Wave III from 2001 to 2002, Wave IV from 2008 to 2009, and Wave V from 2016 to 2018, corresponding to the developmental life stages of adolescence, emerging adulthood, young adulthood, and adulthood.

To account for missing data, we employed multiple imputation using chained equations (Royston and White 2011). To capture data on marital timing, we restricted the sample to those who completed Waves I, II, III, IV, and V of Add Health and Wave I of the Add Health Parent Questionnaire and had valid sample weights. We imputed the following variables: whether the adult child lived with their parents or cohabited at Wave IV, the adult child’s frequency of communication with their mother, the adult children’s marital timing, the adult child’s educational mobility, the adult child’s financial transfers to their mothers in young adulthood, the adult child’s full-time employment status in young adulthood, and the adult child’s income in young adulthood, leading to an analytic sample of 7634 respondents.

5.2. Measures

5.2.1. Dependent Variable

An adult child gave money to their parents in the past year. Our primary dependent variable was financial transfers from adult children to parents, drawn from Wave V of Add Health. It came from the following question: “How many times have you paid a parent or parent figure’s living expenses or given him or her more than USD 50 to pay living expenses during the past 12 months?” Responses included never (1), 1 or 2 times (2), 3 or 4 times (3), and 5 or more times (4). Due to skewness of the outcome variable, we dichotomized responses into the categories of did not give any money to their parent (0) to give any money to their parent (1).

5.2.2. Independent Variable

Marital timing. The primary independent variable of interest was the timing of adult children’s first marriage by Wave V of Add Health, which included the categories of never married (0), married early (1), married on time (2), and married late (3). Adult children who married early were those who married before age 25, adult children who married on time were those who married between ages 25–27, and adult children who married late were those who married at ages 28 and later. These ages were based on categories of marital timing from previous research, which identified early marriage as before age 25 (Uecker and Stokes 2008) and on time marriage, defined as between ages 25–27 (Payne 2015). Thus, marriage at ages 28 and older is defined as late marriage. The timing of marriage was derived from information from the year respondents were first married and the year respondents were born, drawn from Waves I, II, III, IV, and V.

5.2.3. Moderators

Gender. Gender was derived from Wave I of Add Health, whether the respondent identified as a man (1) or woman (2).

Race and ethnicity. Race and ethnicity were a measure created from Wave V of Add Health: whether the respondent identified as white, Black, Hispanic, and Asian.

5.2.4. Control Variables

We included control variables that have been linked to intergenerational financial transfers, including immigrant generation status, age of the child and their parent, number of children, number of siblings, employment status, self-rated health, depressive symptoms, frequency of communication and closeness to mothers, living arrangements, ever cohabiting, educational attainment, educational mobility, income, full-time employment status, and previous exchanges of financial transfers between adult children and their mothers in young adulthood.

Immigrant generation status. Immigration generation status used the question “were you born in the U.S.?” Responses ranged from no (0) to yes (1).

Age at Wave IV. The age of the respondent at Wave IV was generated from the year the respondent was interviewed, subtracted from the year of the respondent’s birth.

Age of parent at Wave I. The age of the respondent’s parent at Wave I was from the following question asked at Wave I of the Add Health Parent Questionnaire: “How old are you?”

Ever cohabited by Wave IV. Ever cohabiting was derived from the question “How many romantic or sexual partners have you ever lived with for one month or more?” Responses were then dichotomized into never cohabited (0) and cohabited with at least one partner (2).

Number of children at Wave I. The number of children was created using the question “How many live births resulted from this pregnancy/these pregnancies?” Responses ranged from 0 to 11.

Number of siblings at Wave IV. The number of siblings was drawn from the question, “How many brothers and sisters do you have, both living and deceased? Include biologically related, adoptive, and stepbrothers or sisters.” Responses range from 0 to 11.

Self-rated health of an adult child at Wave IV. Self-rated health was measured using the question, “How is your general physical health?” Responses ranged from (1) excellent, (2) very good, (3) good, (4) fair, and (5) poor. Responses were then dichotomized from excellent, very good, and good (0) to fair and poor (1) to indicate poor self-rated health.

Depressive symptoms at Wave IV. Depressive symptoms were assessed by using an Add Health CES-D scale (alpha = 0.78) on whether the adult child felt bothered, could not shake off the blues, felt just as good as others (reverse-coded), had trouble keeping mind, felt depressed, felt too tired to do things, and felt sad in the week. Responses ranged from never or rarely (0), sometimes (1), a lot of the time (2), to most of the time or all the time (3).

Frequency of communication with mother at Wave IV. Frequency of communication between the adult child and their parent was measured by using the question, “How often do you and your [mother figure] talk on the telephone, exchange letters, or exchange email?” Responses ranged from never (0), once a year or less (1), a few times a year (2), once or twice a month (3), once or twice a week (4), to almost every day (5).

Maternal closeness at Wave IV. Closeness between the adult child and parent was derived from the question, “How close do you feel to your mother figure?” Responses ranged from not at all close (1), not very close (2), somewhat close (3), quite close (4), to very close (5).

Lived with parents at Wave IV. Co-residence with parents was assessed through a question that asked adult children, “Where do you live now? That is, where do you stay most often?” Responses were consolidated into do not live with parents (0) and live with parents (1).

College degree attainment at Wave IV. College degree attainment was included in the analyses through the question, “What is the highest grade or year of regular school you have completed?” Responses were consolidated into never completed college (0) and completed a college degree or more (1).

Maternal educational mobility at Wave IV. Maternal educational mobility was whether the respondent had completed fewer years of education than their mother (1), the same education level as their mother (2), and more years of education than their mother (3) by Wave IV of Add Health.

Income at Wave IV. Income was from the question, “In {2006/2007/2008}, how much income did you receive from personal earnings before taxes, that is, wages or salaries, including tips, bonuses, and overtime pay, and income from self-employment?” We then categorized responses to less than USD 24,999 (1), USD 25,000–USD 49,999 (2), and USD 50,000 and more (3).

Ever worked full-time at Wave IV. Whether respondents had ever worked full-time by young adulthood was from the following question asked at Wave IV: “Have you ever worked full-time at least 35 h a week at a paying job while you were not primarily a student? Do not include summer work.” Responses included no (0) and yes (1).

Full-time employment status at Wave IV. Full-time employment status was based on the questions, “Are you currently working for pay at least 10 h a week?” and “How many total hours a week do you usually spend at these jobs?” Responses were then categorized into working less than 35 h per week (0) and working more than 35 h per week (1).

Financial transfers from adult children to mothers at Wave IV. Financial transfers from adult children to mothers in young adulthood were from the question, “How many times have you paid your [mother figure’s] living expenses or given her more than USD 50 to pay living expenses during the past 12 months?” Reponses included never (0), yes, number of times unknown (1), yes, 1 or 2 times (2), and yes, 3 or 4 times (3). Responses were then dichotomized to indicate no financial transfers (0) and any financial transfers (1).

5.3. Analytic Approach

We produced summary statistics among the full sample and across marital timing categories, specifically for adult children who had never married, married early, married on time, and married late. We used logistic regression to examine the association between timing of marriage and upward financial transfers given to parents. We present results among the full sample and across gender, as well as race and ethnicity. We used sample weights to ensure that the results were nationally representative of the U.S. population.

6. Results

6.1. Table 1 Summary Statistics of Sample and across Marital Timing

Summary statistics of the full sample and by marital timing status are listed in Table 1. For the full sample, approximately 51% identified as women, 72% identified as non-Hispanic White, 14% identified as non-Hispanic Black, 10% identified as Hispanic, and 4% identified as non-Hispanic Asian. At Wave IV, the average age of respondents was approximately 29 years old, and 4% were born outside of the United States. On average, respondents had approximately one child and about three siblings by Wave IV of Add Health. Most respondents, 80%, were employed full-time at Wave IV of Add Health.

On average, respondents reported 0.55 depressive symptoms, or less than one depressive symptom per week. Adult children communicated with their mothers at least once or twice a week and felt quite close to their mothers. A minority of respondents, 15%, lived with their parents, and a little over a third of respondents completed a college degree by young adulthood (34%). Approximately 30% of respondents had given at least USD 50 or more to their mothers at Wave IV of Add Health. Maternal educational mobility showed that approximately 17% of young adult respondents achieved fewer years of education than their mothers, 41% of young adult respondents achieved the same level of education as their mothers, and 41% of young adult respondents achieved more years of education than their mothers. Approximately 38% of respondents earned less than USD 24,000, 42% earned between USD 25,000 and USD 49,000, and 20% earned USD 50,000 or more at Wave IV.

By Wave V of Add Health, when respondents were in their late thirties, approximately 27% had never married, 36% had married before age 25, 16% had married between 25 and 27 years old, and 21% of respondents had married late at age 28 or later. Approximately 24% of respondents reported that they had provided more than USD 50 to help parents with living expenses in the past year.

Across marital timing categories, there was significant variation in demographic characteristics. Women were significantly less likely to be never married and marry late relative to men and significantly more likely to marry early (58% of women compared to 42% of men) and marry on time (51% of women compared to 49% of men).

Across race and ethnicity, a greater share of white adults married early (78%), married on time (77%), and married late (74%) relative to never marrying (60%). Among Black adults, a greater share of respondents never married (25%), compared to marrying early (8%), marrying on time (10%), and marrying late (13%). Among Asians, a greater share of respondents never married (5%), compared to marrying on time (4%), marrying late (4), and marrying early (3%).

When considering age, those who never married were significantly younger than respondents who married early and married late. Parents of adult children who were never married were significantly younger in age relative to parents of adult children who married early and who married late. Adult children who never married had significantly fewer children than adult children who married early and on time.

Adult children who were never married were significantly less likely to have ever worked full-time at young adulthood (77%) relative to adult children who married on time (82%) and married late (87%).

Adult children who were never married were significantly more likely to report more depressive symptoms than adult children who married early, married on time, and married late.

A greater share of adult children who were never married lived with their parents in young adulthood (30%), relative to adult children who married early (7%), married on time (9%), and married late (13%).

Adult children who never married were significantly less likely to have completed a college degree (27%) relative to those who married early (32%), on time (38%), and married late (43%). Adult children who married on time and married late were more likely to have completed a college degree relative to respondents who married early (32%).

Adult children who were never married were significantly more likely to give money to their mothers in young adulthood (44%), relative to those who married early (24%), married on time (24%), and married late (27%).

Adult children who were never married were also significantly more likely to earn less than USD 24,000 (49%) than adult children who married early (37%), on time (33%), and late (31%) in young adulthood.

Never married adult children were the most likely to give money to their parents in the past year (33%), followed by adult children who married early (22%), adult children who married late (22%), and adult children who married on time (16%). Adult children who married early and married late were significantly more likely to give money to their parents relative to adult children who married on time.

Table 1. Summary statistics of the sample.

Variable	Full Sample	Never Married	Married Early	Married On time	Married Late
Gender (woman)	0.51	0.47	0.58 a	0.51 b	0.45 b,c
Race and ethnicity
NH White	0.72	0.60	0.78 a	0.77 a	0.74 a,b
NH Black	0.14	0.25	0.08 a	0.10 a	0.13 a,b
Hispanic	0.10	0.11	0.11	0.09	0.09
NH Asian	0.04	0.05	0.03 a	0.04	0.04
Age of respondent (Wave IV)	28.71	28.50	28.70 a	28.66	29.05 a,b,c
Age of parent (Wave I)	41.41	41.55	40.78 a	41.36 b	42.35 a,b,c
Immigrant generation status
Foreign-born	0.04	0.05	0.04	0.04	0.05
Number of children (Wave IV)	0.87	0.60	1.24 a	0.80 a,b	0.62 b,c
Ever cohabited (Wave IV)	0.49
Respondent’s number of siblings (Wave IV)	2.69	2.77	2.74	2.55	2.59
Ever worked full-time in young adulthood (Wave IV)	0.96	0.94	0.96 a	0.97 a	0.97 a
Employed full-time (Wave IV)	0.80	0.77	0.78	0.82 a	0.87 a,b,c
Self-rated health of adult child (Wave IV)	0.08	0.12	0.07 a	0.08 a	0.07 a
CES-D Scale (Wave IV)	0.55	0.63	0.54 a	0.50 a	0.50 a,b
Maternal frequency of communication (Wave IV)	4.24	4.21	4.24	4.30	4.22
Maternal closeness of adult child (Wave IV)	4.50	4.50	4.47	4.51	4.55 b
Lived with parents (Wave IV)	0.15	0.30	0.07 a	0.09 a	0.13 a,b,c
Completed a college degree at Wave IV	0.34	0.27	0.32 a	0.38 a,b	0.43 a,b,c
Gave money to mother (Wave IV)	0.30	0.44	0.24 a	0.24 a	0.27 a
Maternal educational mobility (Wave IV)
Less than mother	0.17	0.18	0.18	0.13 a,b	0.18 c
Same as mother	0.41	0.42	0.39	0.44	0.42
More than mother	0.41	0.40	0.43	0.43	0.40
Income (Wave IV)
<USD 24,000	0.38	0.49	0.37 a	0.33 a	0.31 a,b
USD 25,000–USD 49,000	0.42	0.39	0.42	0.45 a	0.43
USD 50,000+	0.20	0.13	0.21 a	0.22 a	0.26 a,b,c
Timing of marriage (Wave V)
Never	0.27	1 b,c,d	0.00	0.00	0.00
Early (24 and younger)	0.36	0.00	1 a,c,d	0.00	0.00
On time (25–27 years old)	0.16	0.00	0.00	1 a,b,d	0.00
Late (Ages 28+)	0.21	0.00	0.00	0.00	1 a,b,c
Child gave money to parent (Wave V)	0.24	0.33	0.22 a	0.16 a,b	0.22 a,c
n	7634	2050	2763	1164	1657

^a—statistically significantly different from never married; ^b—statistically significantly different from married early; ^c—statistically significantly different from married on time; ^d—statistically significantly different from married late. All significantly different at p < 0.05-level or lower.

Table 1. Summary statistics of the sample.

Variable	Full Sample	Never Married	Married Early	Married On time	Married Late
Gender (woman)	0.51	0.47	0.58 a	0.51 b	0.45 b,c
Race and ethnicity
NH White	0.72	0.60	0.78 a	0.77 a	0.74 a,b
NH Black	0.14	0.25	0.08 a	0.10 a	0.13 a,b
Hispanic	0.10	0.11	0.11	0.09	0.09
NH Asian	0.04	0.05	0.03 a	0.04	0.04
Age of respondent (Wave IV)	28.71	28.50	28.70 a	28.66	29.05 a,b,c
Age of parent (Wave I)	41.41	41.55	40.78 a	41.36 b	42.35 a,b,c
Immigrant generation status
Foreign-born	0.04	0.05	0.04	0.04	0.05
Number of children (Wave IV)	0.87	0.60	1.24 a	0.80 a,b	0.62 b,c
Ever cohabited (Wave IV)	0.49
Respondent’s number of siblings (Wave IV)	2.69	2.77	2.74	2.55	2.59
Ever worked full-time in young adulthood (Wave IV)	0.96	0.94	0.96 a	0.97 a	0.97 a
Employed full-time (Wave IV)	0.80	0.77	0.78	0.82 a	0.87 a,b,c
Self-rated health of adult child (Wave IV)	0.08	0.12	0.07 a	0.08 a	0.07 a
CES-D Scale (Wave IV)	0.55	0.63	0.54 a	0.50 a	0.50 a,b
Maternal frequency of communication (Wave IV)	4.24	4.21	4.24	4.30	4.22
Maternal closeness of adult child (Wave IV)	4.50	4.50	4.47	4.51	4.55 b
Lived with parents (Wave IV)	0.15	0.30	0.07 a	0.09 a	0.13 a,b,c
Completed a college degree at Wave IV	0.34	0.27	0.32 a	0.38 a,b	0.43 a,b,c
Gave money to mother (Wave IV)	0.30	0.44	0.24 a	0.24 a	0.27 a
Maternal educational mobility (Wave IV)
Less than mother	0.17	0.18	0.18	0.13 a,b	0.18 c
Same as mother	0.41	0.42	0.39	0.44	0.42
More than mother	0.41	0.40	0.43	0.43	0.40
Income (Wave IV)
<USD 24,000	0.38	0.49	0.37 a	0.33 a	0.31 a,b
USD 25,000–USD 49,000	0.42	0.39	0.42	0.45 a	0.43
USD 50,000+	0.20	0.13	0.21 a	0.22 a	0.26 a,b,c
Timing of marriage (Wave V)
Never	0.27	1 b,c,d	0.00	0.00	0.00
Early (24 and younger)	0.36	0.00	1 a,c,d	0.00	0.00
On time (25–27 years old)	0.16	0.00	0.00	1 a,b,d	0.00
Late (Ages 28+)	0.21	0.00	0.00	0.00	1 a,b,c
Child gave money to parent (Wave V)	0.24	0.33	0.22 a	0.16 a,b	0.22 a,c
n	7634	2050	2763	1164	1657

^a—statistically significantly different from never married; ^b—statistically significantly different from married early; ^c—statistically significantly different from married on time; ^d—statistically significantly different from married late. All significantly different at p < 0.05-level or lower.

6.2. Table 2 Results among Full Sample and by Gender

Analytic results among the full sample and across gender, including all control variables, appear in Table 2. Results are displayed using the reference category of never married. To compare differences across all marital timing groups, results for the reference categories of married early, married on time, and married late are described below, and full results are available upon request.

Table 2. Logistic regressions predicting upward financial transfers to parents based on adult children’s timing of marriage, among full sample and by gender.

	Full Sample		Men		Women
	OR
Timing of an adult child’s marriage (ref. group: Never)
Early	0.82		1.02		0.68	**
On time	0.58	***	0.59	*	0.56	**
Late	0.82		0.92		0.71	*
Gender of adult child (ref. group: Men)	1.04		-		-
Race and ethnicity of respondent (ref. group: White)
Non-Hispanic Black	2.14	***	2.18	***	2.11	***
Hispanic	1.59	**	1.32		1.87	**
Non-Hispanic Asian	2.00	***	1.89	*	2.17	**
Age of adult child (Wave IV)	1.00		0.98		1.03
Age of parent (Wave I)	0.99	*	0.98		0.99
Foreign-born (ref. group: Born in U.S.)	1.73	**	2.15	**	1.43
Adult child’s number of children (Wave IV)	1.07	*	1.13		1.00
Ever cohabited (Wave IV)	1.05		1.04		1.06
Adult child’s number of siblings (Wave IV)	1.02		1.13		1.02
Adult child’s CES-D depressive symptoms (Wave IV)	1.33	**	1.12		1.51	***
Self-rated health of adult child (Wave IV)	1.21		1.36		1.12
Lived with parents (Wave IV)	1.24		1.20		1.28
Adult child completed college degree (Wave IV; ref. group: <College)	0.65	***	0.63	**	0.67	**
Gave money to mother (Wave IV)	3.18	***	3.38	***	3.05	***
Maternal educational mobility (Wave IV; ref. group: <Mother)
Same as mother	1.02		1.11		0.96
More than mother	1.28	*	1.40		1.24
Income (Wave IV; ref. group: <USD 24,000)
USD 25,000–USD49,000	0.90		0.91		0.82
USD 50,000+	0.95		0.89		0.94
Adult child ever worked full-time by young adulthood (Wave IV; ref. group: Never worked full-time)	0.89		0.88		0.86
Adult child worked full-time (Wave IV; ref. group: Never worked)	0.98		1.10		0.91
Frequency of communication with mother (Wave IV)	1.01		1.01		1.01
Maternal closeness (Wave IV)	1.14	*	1.29	**	1.05
Constant	0.07	**	0.07	*	0.06	*
F-statistic	21.38		9.24		14.11
n	7634		3202		4432

Note: OR—odds ratio. * p < 0.05. ** p < 0.01. *** p < 0.001.

Table 2. Logistic regressions predicting upward financial transfers to parents based on adult children’s timing of marriage, among full sample and by gender.

	Full Sample		Men		Women
	OR
Timing of an adult child’s marriage (ref. group: Never)
Early	0.82		1.02		0.68	**
On time	0.58	***	0.59	*	0.56	**
Late	0.82		0.92		0.71	*
Gender of adult child (ref. group: Men)	1.04		-		-
Race and ethnicity of respondent (ref. group: White)
Non-Hispanic Black	2.14	***	2.18	***	2.11	***
Hispanic	1.59	**	1.32		1.87	**
Non-Hispanic Asian	2.00	***	1.89	*	2.17	**
Age of adult child (Wave IV)	1.00		0.98		1.03
Age of parent (Wave I)	0.99	*	0.98		0.99
Foreign-born (ref. group: Born in U.S.)	1.73	**	2.15	**	1.43
Adult child’s number of children (Wave IV)	1.07	*	1.13		1.00
Ever cohabited (Wave IV)	1.05		1.04		1.06
Adult child’s number of siblings (Wave IV)	1.02		1.13		1.02
Adult child’s CES-D depressive symptoms (Wave IV)	1.33	**	1.12		1.51	***
Self-rated health of adult child (Wave IV)	1.21		1.36		1.12
Lived with parents (Wave IV)	1.24		1.20		1.28
Adult child completed college degree (Wave IV; ref. group: <College)	0.65	***	0.63	**	0.67	**
Gave money to mother (Wave IV)	3.18	***	3.38	***	3.05	***
Maternal educational mobility (Wave IV; ref. group: <Mother)
Same as mother	1.02		1.11		0.96
More than mother	1.28	*	1.40		1.24
Income (Wave IV; ref. group: <USD 24,000)
USD 25,000–USD49,000	0.90		0.91		0.82
USD 50,000+	0.95		0.89		0.94
Adult child ever worked full-time by young adulthood (Wave IV; ref. group: Never worked full-time)	0.89		0.88		0.86
Adult child worked full-time (Wave IV; ref. group: Never worked)	0.98		1.10		0.91
Frequency of communication with mother (Wave IV)	1.01		1.01		1.01
Maternal closeness (Wave IV)	1.14	*	1.29	**	1.05
Constant	0.07	**	0.07	*	0.06	*
F-statistic	21.38		9.24		14.11
n	7634		3202		4432

Note: OR—odds ratio. * p < 0.05. ** p < 0.01. *** p < 0.001.

6.3. Results among the Full Sample

Is the timing of adult children’s marriage associated with upward financial transfers from adult children to aging parents? Results among the full sample indicate that relative to adult children who never married, adult children who married on time were significantly less likely to provide upward financial transfers to their parents in adulthood, providing some support for Hypothesis 1a, the argument that marriage continues to operate as a greedy institution.

Relative to adult children who married early (results available upon request), those who married on time were significantly less likely to give upward financial transfers to their adult children, in contrast to Hypothesis 1b, that adult children who married on time would provide more financial transfers than adult children who married early.

Relative to adult children who married on time (results available upon request), adult children who never married, married early, or married late were significantly more likely to provide upward financial transfers to their parents, providing some mixed support for Hypotheses 1a and 1c but not Hypothesis 1b.

Relative to adult children who married late (results available upon request), adult children who married on time were significantly less likely to provide upward financial transfers to parents, providing some support for Hypothesis 1c.

6.4. Results among Men

Our second hypothesis asked whether the association between marital timing and upward financial transfers would hold among men. Results indicate that relative to men who never married, men who married on time were significantly less likely to provide upward financial transfers to their parents, providing partial support for the argument for marriage as a greedy institution. Although men who married early and late were less likely to provide upward financial transfers to parents, these associations were not statistically significant.

Relative to men who married early and men who married late, men who married on time were significantly less likely to provide upward financial transfers to their parents, providing no support that adult children who married early would provide fewer upward financial transfers relative to adult children who married on time, but supported the expectation that men who married late would provide more upward financial transfers than men who married on time. Relative to men who married on time, men who never married, men who married early, and men who married late were significantly more likely to provide upward financial transfers to their parents.

6.5. Results among Women

Among women, relative to those who never married, women who married early, on time, and late were significantly less likely to provide upward financial transfers, offering support for the argument of marriage as a greedy institution among women.

Relative to women who married early, women who married on time, and women who married late, only those who never married were significantly more likely to provide upward financial transfers to parents. Therefore, we did not find evidence to support that adult children marrying late or marrying on time would be associated with differences in levels of upward financial transfers to parents. Nonetheless, these findings also support Hypothesis 2 that marriage operates as a greedier institution for women over men, as married women, regardless of marital timing, were significantly less likely to provide upward financial transfers to their parents.

6.6. Testing Significant Coefficients across Gender Models

To examine whether coefficients across gender models were significantly different among men and women, we included interaction terms of gender and marital timing in models, including the full sample (full results available upon request). In the model including all control variables, we found that women who married early were significantly less likely to provide upward financial transfers to their parents relative to men who married early, providing partial support for Hypothesis 2.

6.7. Supplemental Analyses among Full Sample and across Gender, including Measures of Father Closeness and Frequency of Communication

We included supplemental analyses including measures of father closeness, frequency of communication, and financial transfers in young adulthood, which decreased our sample size from 7634 respondents to 5633 respondents (full results available upon request). In the full model with all controls, relative to adult children who never married, adult children who married on time were significantly less likely to provide upward financial transfers to their parents. Among men, a similar pattern emerged: relative to adult men who never married, adult children who married on time were significantly less likely to provide upward financial transfers to their parents, similar to findings among the full analytic sample. Among women, relative to those who never married, women who married early, on time, and late were significantly less likely to provide upward financial transfers to their parents.

6.8. Table 3 Results by Race and Ethnicity

Analytic results across race and ethnicity, including all control variables, are shown in Table 3. Results are displayed using never married as the reference category. Results examine whether marital timing is associated with upward financial transfers among white, Black, Hispanic, and Asian adults. To compare differences across all marital timing groups, findings for the reference categories of married early, married on time, and married late are also described. Full results are available upon request.

Table 3. Logistic regressions predicting upward financial transfers to parents based on adult children’s timing of marriage by race and ethnicity.

	White		Black		Hispanic		Asian
	OR
Timing of Adult Child’s Marriage (ref. group: Never)
Early	0.65	**	1.10		1.48		0.37	*
On time	0.46	***	0.89		0.72		0.60
Late	0.75		0.87		1.24		0.53
Gender of adult child (ref. group: Men)	0.95		1.20		1.15		1.00
Age of adult child (Wave IV)	0.99		0.99		0.98		1.20
Age of parent (Wave I)	0.99		0.98	*	1.01		1.00
Foreign-born (ref. group: Born in U.S.)	2.12		0.79		1.82	*	0.95
Adult child’s number of children (Wave IV)	1.16	*	0.96		1.05		1.29
Ever cohabited (Wave IV)	1.03		1.27		0.97		1.12
Adult child’s number of siblings (Wave IV)	1.03		1.02		1.01		1.10
Adult child’s CES-D depressive symptoms (Wave IV)	1.37	**	1.29		1.42		0.98
Self-rated health of adult child (Wave IV)	1.72	**	0.63		0.88		0.85
Lived with parents (Wave IV)	1.27		1.19		1.24		1.06
Adult child completed college degree (Wave IV; ref. group: <College)	0.63	**	0.83		0.53	*	1.31
Gave money to mother (Wave IV)	3.65	***	2.50	***	2.26	**	3.89	***
Maternal educational mobility (Wave IV; ref. group: <Mother)
Same as mother	0.93		1.46		0.99		0.46	*
More than mother	1.30		1.24		1.23		1.34
Income (Wave IV; ref. group: <USD 24,000)
USD 25,000–USD 49,000	0.83		0.97		0.77		2.87	*
USD 50,000+	0.81		1.14		1.04		2.12
Adult child ever worked full-time by young adulthood (Wave IV; ref. group: Never worked full-time)	0.82		0.92		1.08		0.76
Adult child worked full-time (Wave IV; ref. group: Never worked)	1.10		1.05		0.77		0.65
Frequency of communication with mother (Wave IV)	1.05		0.94		0.95		1.30
Maternal closeness (Wave IV)	1.21	**	1.19		0.95		0.71
Constant	0.06	*	0.59		0.31		0.00
F-statistic	12.43		2.08		1.91		1.83
N	4725		1456		974		479

Note: OR—odds ratio. * p < 0.05. ** p < 0.01. *** p < 0.001.

Table 3. Logistic regressions predicting upward financial transfers to parents based on adult children’s timing of marriage by race and ethnicity.

	White		Black		Hispanic		Asian
	OR
Timing of Adult Child’s Marriage (ref. group: Never)
Early	0.65	**	1.10		1.48		0.37	*
On time	0.46	***	0.89		0.72		0.60
Late	0.75		0.87		1.24		0.53
Gender of adult child (ref. group: Men)	0.95		1.20		1.15		1.00
Age of adult child (Wave IV)	0.99		0.99		0.98		1.20
Age of parent (Wave I)	0.99		0.98	*	1.01		1.00
Foreign-born (ref. group: Born in U.S.)	2.12		0.79		1.82	*	0.95
Adult child’s number of children (Wave IV)	1.16	*	0.96		1.05		1.29
Ever cohabited (Wave IV)	1.03		1.27		0.97		1.12
Adult child’s number of siblings (Wave IV)	1.03		1.02		1.01		1.10
Adult child’s CES-D depressive symptoms (Wave IV)	1.37	**	1.29		1.42		0.98
Self-rated health of adult child (Wave IV)	1.72	**	0.63		0.88		0.85
Lived with parents (Wave IV)	1.27		1.19		1.24		1.06
Adult child completed college degree (Wave IV; ref. group: <College)	0.63	**	0.83		0.53	*	1.31
Gave money to mother (Wave IV)	3.65	***	2.50	***	2.26	**	3.89	***
Maternal educational mobility (Wave IV; ref. group: <Mother)
Same as mother	0.93		1.46		0.99		0.46	*
More than mother	1.30		1.24		1.23		1.34
Income (Wave IV; ref. group: <USD 24,000)
USD 25,000–USD 49,000	0.83		0.97		0.77		2.87	*
USD 50,000+	0.81		1.14		1.04		2.12
Adult child ever worked full-time by young adulthood (Wave IV; ref. group: Never worked full-time)	0.82		0.92		1.08		0.76
Adult child worked full-time (Wave IV; ref. group: Never worked)	1.10		1.05		0.77		0.65
Frequency of communication with mother (Wave IV)	1.05		0.94		0.95		1.30
Maternal closeness (Wave IV)	1.21	**	1.19		0.95		0.71
Constant	0.06	*	0.59		0.31		0.00
F-statistic	12.43		2.08		1.91		1.83
N	4725		1456		974		479

Note: OR—odds ratio. * p < 0.05. ** p < 0.01. *** p < 0.001.

6.9. Results among White Adults

Among white adults, relative to those who never married by adulthood, those who married early and married on time were significantly less likely to provide upward financial transfers to their parents, providing partial support for the argument of marriage as a greedy institution. Relative to white adults who married early, those who never married were significantly more likely to provide upward financial transfers to their parents. Relative to white adults who married on time, those who never married and married late were significantly more likely to provide upward financial transfers to their parents. Relative to white adults who married late, white adults who married on time were significantly less likely to provide upward financial transfers to their parents.

6.10. Results among Black Adults

Among Black adults, relative to those who never married by adulthood, those who married early and late were more likely to give upward financial transfers to their parents, while those who married on time were less likely to give upward financial transfers to their parents. However, none of these associations were statistically significant. When rotating reference categories of Black adults who married early, married on time, and married late, we did not find any significant associations between adult children’s marital timing and upward financial transfers to parents. In sum, we did not find support for the argument of marriage as a greedy institution among Black adults.

6.11. Results among Hispanic Adults

Hispanic adults who married on time were less likely to provide upward financial transfers to their parents relative to Hispanic adults who were never married. However, none of these associations were statistically significant. In sum, we did not find support for the argument of marriage as a greedy institution among Hispanic adults, comparing Hispanics who were never married to Hispanics who were married.

However, when changing the reference categories, we found that marital timing was significantly associated with upward financial transfers among Hispanic adults. Relative to Hispanic adults who married on time, Hispanic adults who married early were significantly more likely to provide upward financial transfers to their parents.

6.12. Results among Asian Adults

Among Asian adults, relative to those who never married by adulthood, only Asian adults who married early were significantly less likely to provide upward financial transfers to their parents. Relative to Asian adults who married early, we only found that those who never married were significantly more likely to provide upward financial transfers to their aging parents. For other reference categories of Asians who married on time and married late, we did not find any significant associations between adult children’s marital timing and upward financial transfers to parents. Therefore, we found partial support for the argument for marriage as a greedy institution among Asian adults.

6.13. Testing Significant Coefficients across Race and Ethnicity Models

To examine whether coefficients across models were significantly different among racial and ethnic groups, we included interaction terms of race, ethnicity, and marital timing in models including the full sample. Including all control variables, we found that Black adults who married on time were significantly more likely than white adults who married on time to provide upward financial transfers to their parents. We also found that Hispanic adults who married early were significantly more likely than white adults who married early to provide upward financial transfers to their parents. These results align with the prediction that marriage operates as a greedier institution for white adults relative to Black and Hispanic adults, providing support for hypothesis 3.

6.14. Supplemental Analyses across Race and Ethnicity, including Measures of Father Closeness and Frequency of Communication

We included supplemental analyses incorporating measures of father closeness and frequency of communication among respondents who grew up with two parents, including a mother and a father (full results available upon request). Overall, we found that results remained generally consistent with the analytic sample. Among white adults, we found that, relative to those who were never married, white adults who married on time were significantly less likely to provide upward financial transfers to their parents. Among Asian adults, we found that, relative to those who never married, those who married early were significantly less likely to provide upward financial transfers to their parents. Among Black and Hispanic adults, we did not find that marital timing was significantly associated with the likelihood of providing upward financial transfers to parents.

7. Discussion

Prior work has characterized marriage as a “greedy institution” that monopolizes the resources of married adults (Sarkisian and Gerstel 2008). Research has supported this idea by showing that adult children who are married offer less support to their aging parents (Gerstel and Sarkisian 2007; Sarkisian and Gerstel 2008). Nevertheless, prior work has not considered how recent shifts in marital timing may reshape the relationship between an adult child’s marriage and the financial support of their aging parents. We hypothesized that adult children who never married would provide the greatest upward financial support, followed by adult children who married late, adult children who married on time, and adult children who married early. We used data from Waves I–V of Add Health and Wave I of the Add Health Parent Questionnaire to examine whether adult children’s marital timing, specifically whether they were never married, married early, married on time, or married late, was associated with their likelihood of providing upward financial transfers to their aging parents. Overall, we found some support for the argument of marriage as a greedy institution, but substantial variation across gender, race, and ethnicity.

Descriptively, we found that upward financial transfers to parents varied across marital timing categories. The largest proportion of those who provided upward financial transfers to their parents were those who never married (33%), followed by those who married early (22%), those who married late (22%), and those who married on time (16%). Examining our analytic results, we did find support that, relative to those who never married, adults who married on time were significantly less likely to provide upward financial transfers to their parents among the full sample. Among men, only those who married on time were significantly less likely to provide upward financial transfers to their parents relative to men who never married. Among women, those who married early, on time, and late were significantly less likely to provide upward financial transfers relative to women who never married. Among white adults, relative to those who never married, those who married early and on time were significantly less likely to provide upward financial transfers to their parents. Among Asian adults, relative to those who never married, those who married early were significantly less likely to provide upward financial transfers to their parents. Among Hispanic adults, relative to those who married on time, those who married early were significantly more likely to provide upward financial transfers to their parents. However, among Black and Hispanic adults, we did not find that never marrying was significantly associated with greater upward financial transfers from adult children to parents. These results suggest that there is heterogeneity in the role of marriage in intergenerational relationships and that marriage may not necessarily operate as a greedy institution for all groups.

Three key themes emerged from our findings. First, our findings reveal some support that marriage continues to operate as a greedy institution in the context of intergenerational relationships, echoing previous work (Gerstel and Sarkisian 2007; Sarkisian and Gerstel 2008). We found that never married adult children were significantly more likely to transfer financial support to their aging parents than adult children who married on time. We did find some significant differences between marital timing categories. More specifically, we found that adult children who married early and late were significantly more likely to provide upward financial transfers to their parents relative to adult children who married on time. These findings provide mixed support for our predictions. We expected that adult children who married late may be more likely to provide upward financial support to their aging parents relative to individuals who married on time, as late marriage delays accompany competing demands and allow adult children more time to accumulate resources that may also benefit their parents. However, we did not expect that adult children who married early would offer more support to their parents than children who married on time. One possible explanation is that early marriage represents a selective social group. More specifically, adults who marry early may have greater resources that allow them to pursue marriage at earlier ages, which may also benefit their aging parents. Or adults who marry earlier may have certain values, like religious beliefs, that are linked to both early marriage and likelihood of intergenerational support (Silverstein 2006; Uecker 2014). Nonetheless, more research is needed to better understand these findings.

The second theme that emerged was that marriage was a greedier institution for women relative to men, confirming previous studies (Gerstel and Sarkisian 2007; Sarkisian and Gerstel 2008). Among women, those who married early, on time, or late were significantly less likely to transfer financial support to their aging parents relative to never married women. Among men, only those who married on time were significantly less likely to offer financial support to their aging parents relative to men who never married. In addition, women who married early were significantly less likely than men who married early to provide upward financial transfers to their parents. Why might we expect that marriage is a greedier institution among women relative to men? Marriage may use more of women’s resources than men’s resources. Women often invest more time in labor related to marriage, including housework and childcare (Bianchi et al. 2012). Trends in the gendered division of labor within marriages may spillover into the financial resources offered to one’s aging parents for married women versus married men, with married women offering less support to aging parents than married men. Moreover, due to gender discrimination in the labor market (Gharehgozli and Atal 2020), women may already have more limited and fewer financial resources than men and therefore have fewer financial resources to give back to their aging parents.

Finally, the third theme of our study points toward notable differences in the relationship between marital timing and upward financial transfers by racial and ethnic background. These findings suggest that marriage operates as a greedy institution amongst white and Hispanic adults who married on time and Asian adults who married early. Among white adults, those who married early or on time were significantly less likely to transfer financial resources to their aging parents than white adults who never married. These findings suggest that, generally, marriage operates as a greedy institution among white adults, depending on the timing of their marriage. In addition, relative to white adults who married on time, white adults who married early and married late were significantly more likely to provide upward financial transfers to their parents.

We predicted that marriage would operate as a greedier institution among Black, Hispanic, and Asian American adults given stronger norms to provide support across generations (Lanuza 2020). Our results partially support these expectations. Among Hispanic adults, relative to those who married on time, those who married early were significantly more likely to provide upward financial transfers to their parents. Although these findings suggest that marital timing is linked to upward financial transfers to parents, Hispanic adults who married early still provided significantly more support to their aging parents than their white peers who married early. Potential explanations for this include a greater sense of familism and obligations to provide financial support to parents due to sacrifices made during immigration among Hispanic adults relative to white adults (Lanuza 2020; Silverstein 2006). Thus, although marriage may be linked to upward financial transfers among Hispanic adults for some timing categories, marriage appears to be less greedy for Hispanic adults than white adults.

Among Asian adults, we found that, relative to those who never married, adult children who married early were significantly less likely to provide upward financial transfers to their parents. This suggests that marriage may operate as a greedy institution among Asian American adults, but only for those who marry early, who may have fewer resources to share with their aging parents. We did not find that the results for Asian American adults differed significantly from white adults, and thus, we cannot conclude whether marriage is a greedier institution for white adults relative to Asian American adults. Among Black adults, however, neither marriage nor marital timing were significantly associated with upward financial transfers to aging parents, suggesting that marriage may not operate as a greedy institution among Black adults. In addition, we found that Black adults who married on time were significantly more likely than white adults who married on time to provide upward financial transfers to their parents. It is possible that among Black adults, due to closer and stronger family ties, there is a greater expectation of giving back to parents than white adults, including financial, emotional, and instrumental support, regardless of marital status (Lanuza 2020; O’Brien 2012; Zhang and Grant 2023).

Although our work adds to a growing body of literature focused on how adult children’s role acquisition impacts parental well-being and intergenerational relationships (Zhang and Hammersmith 2023), there are several limitations to acknowledge. First, our marital timing categories were based on average timing of marriage for the cohort of adults who came of age into young adulthood around 2008–2009, which may reflect earlier ages at first marriage than today. Moreover, we use the same marital timing categories across all groups, which may not necessarily reflect more specific patterns of differences in marital timing across gender, race, and ethnicity. We also focus exclusively on upward financial transfers to measure intergenerational exchange, as adult children in the Add Health Study do not report on transfers of time or emotional support to aging parents. Third, the upward financial transfer measure is a broad indicator of whether the adult child provided at least USD 50 to help their parent with their living expenses in the last year, which obscures the magnitude and frequency of financial support offered to aging parents. Add Health did not contain other measures of upward financial transfers to parents. However, this measure may be a proxy for parents who have greater financial need, as it is money towards parents’ living expenses and represents a smaller dollar amount. Fourth, we do not have data on parents’ socioeconomic status beyond measures included in Wave I of Add Health. Thus, we are unable to assess the degree to which parents may need financial assistance from their adult children in later life. Last, dyadic studies show that parent and child reports are not necessarily concordant (Kenny 2011). Since we only used data from the adult child for the measure of financial transfers, we cannot confirm whether parents would agree with adult children’s reported levels of financial support.

Our work also provides direction for additional research. Future work should consider the role of adult children’s cohabitation on upward financial transfers, as many adult children delay or forego marriage in favor of living with their partner (Manning 2013). Future work should consider whether cohabitation operates as a similarly “greedy institution” in the context of intergenerational relationships. Finally, future work can use data from the Health and Retirement Study, as well as Wave II of the Add Health Parent Study, to examine how adult children’s marital timing may shape other intergenerational exchanges, such as instrumental and caregiving support.

Despite these limitations, this study offers contributions to current understandings of how the timing of children’s entry into adult roles shapes support offered to aging parents. Prior work has already demonstrated that adult children’s entry into marriage jeopardizes support directed toward aging parents (Gerstel and Sarkisian 2007; Sarkisian and Gerstel 2008), yet we know little about how historic delays in adult children’s entry into marriage may reshape the financial support adult children are able to offer their aging parents. Our results suggest that marital timing matters, and marriage persists as a greedy institution, especially among men, women, white, Hispanic, and Asian adults. This may be of concern to policymakers and practitioners, as aging parents may increasingly rely on support from adult children, who may not be positioned to provide necessary support. Support deficits may become more noticeable as aging parents live longer, while not always in great health (Boersma 2020; Jindai 2016), accompanied by gaps in access to other social safety nets intended to support older Americans, such as Medicare, Medicaid, Social Security, and other forms of social welfare (Freedman et al. 2024).