Determinants of Living in a Three-Generation Household among Adolescents of Ethnic Groups in the U.S.: Family Structure, Social–Economic Status, and Cultural Factors

Abstract

Purpose: Multigenerational families are becoming increasingly common in the U.S. This trend is primarily driven by three-generation households with grandparents. The coresident grandparents play an important role in adolescents’ health and well-being. Thus, by focusing on three-generational households, this study examined the determinants of living in three-generational households among adolescents within the contexts of the social–economic, cultural, and family factors that influence grandparent co-residence by ethnic groups. Methods: This study used the National Longitudinal Study of Adolescent to Adult Health (Wave I–III). The study sample included 10,093 adolescents, including White, African American, Asian, and Hispanic youth. This study conducted a series of logistic regression models to examine the associations between co-residence with grandparents and significant predictors in family structure, socioeconomic status, and cultural factors for youths in the U.S. by ethnic groups. Results: For White families, lower socioeconomic status was more pertinent to three-generational co-residence. However, the associations were in the opposite direction for Hispanic and African American households, indicating that higher socioeconomic status families were found to live with grandparents in those groups. For Hispanic families, adolescents from Spanish-speaking homes were more likely to live in three-generational households than those from English-speaking homes. Implications: These results suggest that family characteristics in three-generational households vary by ethnic group. Notably, family cultural factors were significant determinants of co-residence with grandparents in three-generational households, especially in Hispanic families. This study contributes to the sustainability discourse by examining the intersectionality of cultural maintenance, health and well-being, and aging society among three-generational households in the United States.

1. Introduction

In the United States (U.S.), the nuclear family structure of two parents and their children has long been considered the normative family structure. However, it has become less prevalent. The proportion of nuclear families dropped to 19% in 2013 from 40% in 1970 [1], multigenerational families have become more common [2]. As defined by the United States Census Bureau [3], multigenerational families have three or more generations of parents and children. Over the past 40 years, the percentage of Americans living in multigenerational homes has more than doubled, from 7% in 1971 to 18% in 2021 [4]. This trend is driven mainly by three-generation households with grandparents. A three-generational household has a significant impact on the health and well-being of adolescents [5]. Therefore, it is essential to gain insight into the dynamics of a three-generational household. The rise in three-generational households is attributed to three factors: family structure, social–economic factors, and cultural context. First, demographic changes in the U.S. have increased the number of three-generational households. With increased life expectancies, people are living longer and can spend more time with their children and grandchildren creating a stronger bond between generations. This has led to a resurgence in three-generational families, where grandparents, parents, and children all live together [6,7]. Having non-marital children and divorce are also associated with a higher likelihood of co-residence with grandparents [8]. Three-generation households are more common among children with single parents than in two-parent households [8]. In 2009, there were 7.8 million children living with a grandparent, 13% (1.0 million) also living with two parents, and 33% (2.6 million) living with a single parent [8]. Families with only one parent tend to have social and economic instability, resulting in grandparents caring for their grandchildren more often than two-parent families [9,10].

Second, social–economic status should also be considered when examining the effects of family characteristics on co-resident grandparents. The economic stagnation that occurred during the recession of 2007–2009 may have contributed to the growth of three-generational households [7,11,12]. According to the Pew Research Center, the number of children living with a grandparent increased rapidly from 2007 to 2008 as a result of the recession [13]. People with fewer financial resources may seek assistance from three-generational families. It is likely that many parents, children, and grandparents lived together to reduce housing costs and facilitate resource-sharing between generations since households often depend on informal social supports (such as family members and the local community) when faced with external financial constraints. Three-generational households likely also increased the grandparents’ role in caring for their grandchildren [14,15].

In addition, cultural factors may also be significant determinants of co-residence with grandparents. Asians and Hispanic immigrants to the United States have contributed to the increase in the number of three-generational households due to their tendency to co-reside with grandparents more than native-born families [13]. In comparison to non-Hispanic White families (7%), Hispanic (14%), Asian (14%), and Black (17%) families are more likely to live in three-generational households [8]. This phenomenon may be associated with socioeconomic inequalities, but scholars also suggest that it relates to Hispanic and Asian cultural values that emphasize the responsibilities and roles of families [16]. For example, familism is an important cultural value among Hispanics [17]. In Asian culture, filial piety is one of the most important cultural values [18]. These values include respecting and caring for elders, maintaining strong family ties, and facilitating intergenerational socialization, which may explain why these families are more likely to live together in three-generational households [17,18]. Although three-generational households are on the rise, previous studies mainly described the living arrangements of older adults living alone or living with others, overlooking research on three-generational households [19]. There has also been a limited understanding of how three-generational households are formed.

Therefore, using a nationally representative dataset, The National Longitudinal Study of Adolescent to Adult Health (Add Health), this study aimed to investigate the factors that determine adolescents’ living arrangements in three-generational households, specifically focusing on the influence of socioeconomic, cultural, and family factors on grandparent co-residence within different ethnic groups.

2. Methods

2.1. Data

This study used data from the Wave I (1994–1995), Wave II (1996), and Wave III (2001) surveys of the National Longitudinal Study of Adolescent and Adult Health (Add Health), which followed a nationally representative sample of 20,745 adolescents in grades 7 through 12 in the U.S. since 1994. Add Health used a multistage, stratified, and school-based cluster sampling method. Using the unequal probability of selection, 80 high schools and 52 middle schools were selected in 1994, and then all students in the selected schools were interviewed [20]. After the first in-school survey, adolescents were identified with the unequal probability of selection from these schools and participated in follow-up, in-home surveys during Waves II and III. Chinese, Cuban, and Puerto Rican adolescents were oversampled, and other ethnic groups were sampled in proportion to their actual sizes within the U.S. population [21].

2.2. Sample

The appropriate sampling weight was employed to merge the dataset from Wave I to Wave III for the longitudinal analysis [20]. After merging the dataset with sampling weights used for longitudinal analysis [20], the number of participants was 10,828 who were enrolled in grades 7–11 during 1994–1995 (Wave I) and interviewed in all three waves (1995, 1996, and 2001). The analyses were further restricted to adolescents from non-Hispanic White, non-Hispanic African American, Asian, and Hispanic backgrounds. A total of 290 adolescents who identified themselves as Native American, multiracial, or of an unlisted race/ethnicity and 445 adolescents living without parents were excluded. The analytic sample consisted of 10,093 adolescents. All analyses were performed using STATA, version 14.0. All results were weighted for national representation, and standard errors were corrected to control for survey design effects of clustering sampling.

2.3. Variables

2.3.1. Dependent Variable

Co-residence with grandparents. Co-residence with grandparents was constructed based on the information from detailed household rosters in which adolescents in Waves I–III reported on the relationship of every household member. Co-residence with grandparents was coded 0 for adolescents under 19 years old who did not have an experience of living with a grandparent and 1 for adolescents under 19 years old who had an experience living with a grandparent.

2.3.2. Independent Variables

Family structure. Using the information from detailed household rosters in which adolescents in Wave I reported on their relationship to every household member, household type was coded 0 for adolescents who lived in single-parent households and 1 for adolescents who lived in two-parent households, regardless of the presence of grandparents.

Socioeconomic factors. To measure family socioeconomic characteristics, family income, parental education, and maternal employment status were included in Wave I. Family income was log-transformed to account for skewed data. Approximately 23% of family income was missing cases in Wave I. It was separately imputed using two supplemental variables (e.g., the median household income of the respondent’s census block and the mean parent’s income of the respondent’s school) and merged into the imputed data to increase the accuracy of imputation [22]. The highest level of education of the two parents in the household was used as an indicator of parental education (less than high school, completed high school or GED, and some college through advanced degree(s)). Maternal employment was coded as 1 if the respondent’s resident mother was not employed at Wave I, coded as 2 if she worked or had a job, and coded as 3 if there was no mother in the household.

Cultural factors. Adolescent immigrant generation, language spoken at home, family cohesion, and parent–child relationship were included to measure the family cultural factors in a household. The immigrant generation combined the measures of nativity (birthplace) and the birthplace of the parents. First generation refers to adolescents who were born outside of the U.S. with at least one foreign-born parent. Second generation refers to adolescents who were born in the U.S. with at least one foreign-born parent, and third generation refers to adolescents who were born in the U.S. with two US-born parents. Language spoken at home was the self-reported primary language spoken at home and categorized as English, Spanish, and other. Family cohesion was measured as the mean of five items (α = 0.75). Responses ranged from 1 (“not at all”) to 5 (“very much”). The items included: (1) “How much do you feel that your parents care about you?” (2) “How much do you feel that people in your family understand you” (3) “How much do you feel that you and your family have fun together?” (4) “How much do you feel that your family pays attention to you?” (5) “How much do you feel that you want to leave home?” Higher scores indicated higher levels of family cohesion. The parent–child relationship was measured as the mean of four items (α = 0.72), ranging from 1 (“Not at all”) to 5 (“very much”). The items included: (1) “How close do you feel to your mother?” (2) “How much do you think your mother cares about you?” (3) “How close do you feel to your father?” (4) “How much do you think your father cares about you?” Higher scores indicated a closer relationship with their parents.

2.3.3. Control Variables

Adolescent demographic characteristics. Adolescent demographic characteristics included age, gender, and race/ethnicity at Wave I. Age was squared to account for nonlinearity in the effect of age. Gender was coded 0 for male and 1 for female. The adolescent’s ethnicity was self-reported at Wave I and categorized as Hispanic, non-Hispanic White, non-Hispanic African American, or Asian American.

Co-resident grandparent characteristics. The number of grandparents at home, duration of stay at home, and age of both grandparents were included to measure the co-resident grandparents’ characteristics. The number of grandparents was categorized as only grandfather, only grandmother, and both. Duration of stay at home was measured as the mean time in years of two grandparents if both lived in the household.

2.3.4. Analytic Strategy

First, a descriptive analysis was conducted to understand general characteristics by racial/ethnic group. Second, stepwise logistic regression models were used to examine the associations between co-residence with grandparents and significant predictors in family contexts (e.g., family structure, socioeconomic factors, and cultural factors) so that the unique contributions of each factor could be evaluated. Model 1 examined the relationships between the presence of grandparents in the household and the family structure. Model 2 added family socioeconomic factors, and Model 3 added family cultural factors and the quality of family relationships. Two family cultural factors (e.g., immigrant status and language at home) were only applied to Asian and Hispanic families due to small statistical power issues among White and Black adolescents. Then, the full logistic regression model (Model 4) was stratified by racial/ethnic groups (i.e., non-Hispanic White, non-Hispanic Black, Hispanic, and Asian).

3. Results

Table 1. Weighted sample descriptive by race/ethnicity (N = 10,093).

	White		African American		Hispanic		Asian		Total
	Mean or %	S.D.	Mean or %	S.D.	Mean or %	S.D.	Mean or %	S.D.	Mean or %	S.D.
Co-residence with grandparents (%)
Living with a grandparent	4.11	0.00	10.54	0.01	9.36	0.01	14.01	0.02	7.61	(0.00)
Not living with a grandparent	95.89	0.00	89.46	0.01	90.64	0.01	85.99	0.02	92.39	(0.00)
Family Structure (%)
Single parent households	19.98	0.01	52.61	0.02	23.50	0.02	15.76	0.02	25.19	(0.01)
Two-parent households	80.02	0.01	47.39	0.02	76.50	0.02	84.24	0.02	74.81	(0.01)
Adolescent demographic characteristics
Age	14.95	0.13	15.14	0.19	15.08	0.22	15.18	0.28	15.01	(0.11)
Gender (%)
Male	49.90	0.01	48.96	0.02	51.79	0.02	53.92	0.03	50.16	(0.01)
Female	50.10	0.01	51.04	0.02	48.21	0.02	46.08	0.03	49.84	(0.01)
Grandparents characteristics
Number of grandparents (%)
Only grandfather	15.14	0.03	5.54	0.02	21.30	0.06	9.80	0.05	13.25	(0.02)
Only grandmother	68.52	0.04	75.37	0.06	50.63	0.07	65.24	0.07	66.62	(0.03)
Both	16.34	0.03	19.09	0.05	28.07	0.05	24.96	0.07	20.13	(0.03)
Length of stay at home	4.96	0.41	4.13	0.49	4.88	0.54	5.02	0.63	4.79	(0.28)
Age of grandfather	68.89	1.46	67.97	2.82	65.13	1.48	70.87	2.50	67.80	(1.01)
Age of grandmother	68.47	1.00	62.80	0.74	66.46	1.50	70.54	1.97	66.67	(0.67)
Family socioeconomic characteristics
Family Income (logged)	10.56	0.05	9.99	0.07	10.16	0.07	10.59	0.06	10.42	(0.05)
Parent Education (%)
Less than high school	7.41	0.01	13.27	0.02	35.70	0.04	13.82	0.03	12.06	(0.01)
Completed high school or GED	29.88	0.02	39.28	0.03	29.01	0.02	21.53	0.03	30.85	(0.01)
Some college through advanced degree(s)	62.71	0.02	47.45	0.04	35.29	0.03	64.65	0.04	57.08	(0.02)
Maternal employment (%)
Employed	84.81	0.01	78.51	0.02	72.99	0.03	86.63	0.02	82.47	(0.01)
Non-employed	11.20	0.01	17.11	0.02	22.62	0.02	8.61	0.01	13.41	(0.01)
No mother in household	3.99	0.00	4.38	0.01	4.40	0.01	4.76	0.01	4.13	(0.00)
Family cultural characteristics
Adolescent immigrants (%)
1st generation	0.66	0.00	1.16	0.00	22.31	0.03	46.33	0.06	5.29	(0.01)
2nd and 3rd generations	99.34	0.00	98.84	0.00	77.69	0.03	53.67	0.06	94.71	(0.01)
Language at home (%)
English	99.53	0.00	99.77	0.00	57.09	0.04	59.19	0.06	92.65	(0.01)
Spanish	0.00	0.00	0.06	0.00	41.31	0.04	0.00	0.00	5.12	(0.01)
Others	0.47	0.00	0.17	0.00	1.60	0.01	40.81	0.06	2.23	(0.00)
Family cohesion	4.02	0.02	4.01	0.03	4.05	0.04	3.95	0.05	4.02	(0.02)
Parent–child relationship	4.63	0.01	4.71	0.02	4.65	0.02	4.57	0.04	4.65	(0.01)
N	5574		2101		1662		756		10,093

Note. Standard deviation is in parentheses; numbers in cells are weighted percentages unless otherwise noted.

presents weighted descriptive characteristics from the analytic sample by racial/ethnic groups. Asian adolescents had the highest proportion among all groups to have lived with a grandparent (14.01%), whereas White adolescents had the lowest rate of living with a grandparent (4.11%). With regard to family structure, African American adolescents had the highest rate of living in single-parent households (52.61%), whereas Asian adolescents had the lowest proportion of living in single-parent households (15.76%). On average, adolescents in three-generational households had lived with a grandparent for 4.79 years. The average age of the grandparents was around 66 years for grandmothers and 67 years for grandfathers. With regard to family cultural characteristics, approximately 22% of Hispanic and 46% of Asian adolescents were first-generation immigrants, whereas nearly all White and African American adolescents were born in the U.S. Approximately 40% of Hispanic and Asian adolescents spoke Spanish or another language as their primary language at home.

Table 2. Logistic regressions of family characteristics on co-residence with grandparents (N = 10,093).

	Step 1		Step 2		Step 3
	Odds	S.E.	Odds	S.E.	Odds	S.E.
Family Structure(ref. single-parent households)
Two-parent households	0.32 ***	(0.12)	0.33 ***	(0.12)	0.33 ***	(0.12)
Adolescent demographic characteristics
Gender (ref. male)
Female	1.30 *	(0.11)	1.30 *	(0.11)	1.33 *	(0.11)
Race (ref. White)
African American	1.89 ***	(0.12)	1.85 ***	(0.12)	1.81 ***	(0.20)
Hispanic	2.35 ***	(0.17)	2.39 ***	(0.19)	2.01 ***	(0.20)
Asian	4.24 ***	(0.23)	4.41 ***	(0.23)	3.65 ***	(0.31)
Age (squared)	1.00	(0.00)	1.00	(0.00)	1.00	(0.00)
Family socioeconomic characteristics
Family Income (logged)			1.00	(0.03)	1.00	(0.03)
Parent Education (ref. less than high school)
Completed high school or GED			1.52 *	(0.21)	1.65 *	(0.23)
Some college through advanced degree(s)			1.23	(0.21)	1.35	(0.23)
Maternal employment (ref. non-employed)
Employed			0.71 *	(0.16)	0.70 *	(0.16)
No mother in household			0.82	(0.27)	0.83	(0.27)
Family cultural factors
Adolescent immigrants (ref. 2nd and 3rd generations)
1st generation					0.9	(0.3)
Language at home (ref. English)
Hispanic					1.65	(0.38)
Others					1.85	(0.44)
Family cohesion					0.93	(0.10)
Parent–child relationship					1.34 *	(0.13)
Constant	0.13	(0.33)	0.12	(0.53)	0.04	(0.78)
N	10,093		10,093		10,093

Note. Standard errors are in parentheses; * p < 0.05 and *** p < 0.001.

presents the odds ratios of family characteristics on co-residence with grandparents from stepped regression models. In Step 1, controlling for family structure and adolescents’ demographic characteristics, adolescents in two-parent households were less likely to live with a grandparent ([OR] = 0.32, p < 0.001) compared to those in single-parent households. African American ([OR] = 1.89, p < 0.001), Hispanic ([OR] = 2.35, p < 0.001), and Asian ([OR] = 4.24, p < 0.001) adolescents were more likely to live with a grandparent compared to White adolescents. Controlling for family socioeconomic characteristics, the significant relationships between family structure, race/ethnicity, and co-residence with grandparents were maintained.

The results for family socioeconomic characteristics (Step 2 of Table 2) indicated that adolescents who had parents with completed high school or GED educations showed a higher likelihood of co-residence with a grandparent ([OR] = 1.52, p < 0.05) compared to those with parents with less than a high school education. In contrast, adolescents whose mothers were employed had lower odds of living with a grandparent ([OR] = 0.70, p < 0.05) compared to those whose mothers were unemployed. Step 3 controls for the family structure, demographic, and socioeconomic variables from Step 2, along with family cultural factors. Overall, the results were similar to those in Step 2. Additionally, among family cultural factors, adolescents who had a closer relationship with their parents were more likely to live with a grandparent ([OR] = 1.34, p < 0.05). However, the findings showed that generational status, language at home, and family cohesion were not significantly associated with co-residence with grandparents after adjustment.

Table 3. Logistic regressions of family characteristics on co-residence with grandparents by race/ethnicity.

	White		African American		Hispanic		Asian
	OR	(SE)	OR	(SE)	OR	(SE)	OR	(SE)
Family Structure (ref. single-parent households)
Two-parent house holds	0.33 ***	(0.18)	0.16 ***	(0.04)	0.53	(0.34)	0.46	(0.57)
Adolescent demographic characteristics
Gender (ref. male)
Female	1.36	(0.17)	1.71 *	(0.43)	1.16	(0.26)	1.02	(0.29)
Age (squared)	1.00	(0.00)	0.99 *	(0.00)	1.00	(0.00)	1.00	(0.00)
Family socioeconomic characteristics
Logged Family Income	0.97	(0.05)	0.95	(0.05)	1.14	(0.10)	1.28	(0.20)
Parent Education (ref. less than high school)
Completed high school or GED	0.93	(0.33)	2.37 *	(0.88)	3.15 *	(0.40)	1.13	(0.61)
Some college through advanced degree (s)	0.76	(0.34)	2.34 *	(0.91)	1.80	(0.34)	0.97	(0.53)
Maternal employment (ref. non-employed)
Employed	0.59 *	(0.24)	0.68	(0.17)	0.87	(0.38)	1.94	(0.59)
No mother in the household	0.69	(0.39)	1.15	(0.62)	0.86	(0.47)	1.61	(0.74)
Family cultural factors
Adolescent immigrants (ref. 2nd and 3rd generations)
1st generation	--	--	--	--	0.58	(0.42)	1.01	(0.56)
Language at home (ref. English)
Spanish	--	--	--	--	2.26 *	(0.40)	--	--
Others	--	--	--	--	1.94	(0.93)	1.17	(0.50)
Family cohesion	1.11	(0.14)	0.69	(0.13)	0.90	(0.27)	1.06	(0.23)
Parent–child relationship	1.12	(0.18)	1.88 *	(0.46)	1.65	(0.30)	0.90	(0.26)
Constant	0.06	(1.09)	0.14	(1.45)	0.01	(1.70)	0.01	(2.95)
N	5570		2095		1661		755

Note. Standard errors are in parentheses; * p < 0.05 and *** p < 0.001.

shows the results of logic regressions to examine whether the associations between family characteristics and grandparent co-residence varied by adolescent race/ethnicity. In terms of family structure, White and African American adolescents in two-parent households were less likely to live with a grandparent compared to those in single-parent households ([OR] = 0.33, p < 0.001 and [OR] = 0.16, p < 0.001, respectively). With regard to family socioeconomic characteristics, African Americans whose parents completed high school, a GED, or some college through advanced degree(s) showed a higher probability of living with a grandparent compared to those whose mothers completed less than a high school degree ([OR] = 2.37, p < 0.05 and [OR] = 2.34, p < 0.05, respectively). Similarly, Hispanic adolescents with parents who completed high school or a GED were more likely to live with a grandparent ([OR] = 3.15, p < 0.05). On the contrary, White adolescents whose mothers were employed had lower odds of living with a grandparent compared to those whose mothers were unemployed ([OR] = 0.59, p < 0.05). Interestingly, with regard to family cultural factors, among Hispanic adolescents, adolescents who spoke Spanish as their primary language at home had a higher probability of living with a grandparent compared to those who spoke English as their primary language at home ([OR] = 2.26, p < 0.05). However, among Asian adolescents, none of the factors were associated with living with a grandparent.

In terms of the quality of family relationships, among Black adolescents, families with a closer relationship between parents and children showed a higher probability of co-residence with grandparents ([OR] = 1.88, p < 0.05).

4. Discussion

Using nationally representative longitudinal data, this study examined the family contextual factors for co-residence with a grandparent in three-generational households and how these relationships differ by race/ethnicity. Overall, family characteristics such as family structure, socioeconomic factors, and cultural factors were found to be related to co-residence with grandparents, and the relationships varied across racial/ethnic groups. Several key findings emerged. First, adolescents in single-parent households were more likely to live with a grandparent than those in two-parent households. Additionally, socioeconomic factors were associated with co-residence with a grandparent, but the direction differed by race/ethnicity. Previous research has found that, in general, families with lower levels of socioeconomic status are more likely to live in three-generational households [12,23]. In the current study, this pattern applied only to White adolescents living in three-generational households. For example, White adolescents whose parents had lower levels of educational achievement were less likely to live with a grandparent. On the contrary, for Hispanic and African American adolescents, this study found that adolescents whose parents had higher education levels were more likely to live in three-generational households. Indeed, lower socioeconomic status was more pertinent to three-generational co-residence among White households, but the associations were in the opposite direction for Hispanic and African American households, indicating that higher socioeconomic status families were found to live with grandparents in those groups. This study also found that Hispanic and Asian adolescents’ co-residence with a grandparent was not significantly related to family structure. As a result of these findings, it can be inferred that economic strain is not the main factor behind living with grandparents, especially in Hispanic and Asian families.

Another key finding is that Hispanic family culture was associated with a higher likelihood of co-residence with a grandparent. Even after controlling for family socioeconomic factors, as expected, the results showed that Hispanic adolescents who spoke Spanish as their primary language at home were more likely to live with a grandparent than those who spoke English as their primary language. Family cultural values and the degree of acculturation are closely related to the primary language used at home [24]. Although this study was not able to confirm the causal effect of primary language at home on three-generational co-residence, it is possible that grandparent co-residence reinforces grandchildren’s language heritage or that households with Hispanic adolescents who speak Spanish at home maintain their family-based cultural orientation. Hispanic families are characterized by family-oriented cultural orientations that value maintaining family relationships across multiple generations. It is expected that families foster strong bonds of attachment and support their children [17]. As a result, strong obligations to support family members are more important to Hispanics [16,25,26]. The social expectation and responsibility of familism establish a mindset of reciprocal duty between grandparents and adult children [11]. Families speaking Spanish at home may have a greater tendency to retain their cultural characteristics, including bringing elderly family members into the home to help their adult children and grandchildren [27,28]. The sustainability of cultural maintenance for immigrants of Hispanic origin and their health and well-being can have important social and economic benefits, such as enhancing cultural diversity, promoting healthy aging, and reducing healthcare costs. These results also highlight the importance of maintaining cultural traditions and practices within the family unit, which can contribute to the sustainability of cultural identity and heritage for the immigrant population.

For Black adolescents, the positive association between the parent–child relationship and co-residence with grandparents can have significant implications for family dynamics and well-being. This association can be explained by grandparents’ unique role in supporting and strengthening the parent–child relationship. When grandparents live in the same home as their children and grandchildren, they provide additional sources of support and association, which can help reduce stress and increase confidence in parenting abilities [29].

To sum up, these findings suggest that for White adolescents, the low socioeconomic status of families was a contributing factor, and for Black adolescents, the parent–child relationship was an important factor for co-residence with grandparents. For Hispanic adolescents, three-generational co-residence can be explained by the cultural orientations of familism and the high socioeconomic status of families. Even though Asian adolescents had the highest proportion of co-residence with a grandparent compared to other racial/ethnic groups, this study did not find any significant cultural factors related to grandparent co-residence for Asian adolescents. One possible explanation is that the relatively small sample size might undermine the statistical power and reduce the chances of finding statistically significant results. In addition, Asian adolescents may be a more heterogeneous and diverse group relative to language and family culture than other racial/ethnic categories.

This study has several limitations. First, although this study used several statistical and empirical approaches to address selection bias, the estimates could be biased if any other significant observed or unobserved variables were omitted from the models. For example, while this study included a rich set of covariates, this study was not able to control for the perceptions and attitudes about grandparents and roles of grandparents, which possibly might be associated with co-residence with grandparents because Add Health did not collect these variables. Second, this study was not able to consider substantial heterogeneity within racial/ethnic categories due to small sample sizes.

Despite these limitations, this study makes some important contributions to the study of determinant factors for grandparent co-residence in three-generational households. This study places this understanding within family contexts in an attempt to examine these relationships in a way that is more relevant to family characteristics and socioeconomic and family cultural factors. Additionally, this study is unique because it compares family characteristics of three-generational households by racial/ethnic groups. To clarify the process underlying the racial/ethnic differences observed in this study, it is critical to consider the role of grandparents and the type of household formation (e.g., host versus guest) in future studies. This study is unique in comparing family characteristics of three-generational households to multiple groups by family structure and race/ethnicity. In future studies, it is important to consider the role of grandparents and the formation type of three-generational households (e.g., host versus guest) to clarify the process underlying the racial/ethnic differences observed in this study. Future studies also should explore variations in sub-group cultures and examine the effects of co-residence with grandparents on the physical and mental well-being of their grandchildren [5]. Research highlights that determinants of three-generational co-residence vary across racial/ethnic groups. Thus, policy implications should be multifaceted to consider the complexity of circumstances and should provide effective resources to meet the varying needs of three-generational households. In recent years, there has been much discussion about modifying the social safety net as the elderly population grows [30]. These policy discussions should focus not only on grandparents but also on the grandchildren with whom they live.

The aging and well-being of both grandparents and grandchildren in three-generational households have significant implications for the development of sustainable policies and programs aimed at supporting aging populations. By understanding the dynamics and needs within these households, policymakers can implement effective measures to promote the overall welfare and quality of life for all generations involved while ensuring the long-term sustainability of support systems for aging populations.