Fictive Kin Networks among African Americans, Black Caribbeans, and Non-Latino Whites

Abstract

Using data from the National Survey of American Life, we investigated the social and demographic correlates of fictive kin network involvement among African Americans, Black Caribbeans, and non-Latino Whites. Specifically, we examined the factors shaping whether respondents have fictive kin, the number of fictive present kin in their networks, and the frequency with which they received support from fictive kin. Eighty-seven percent of respondents had a fictive kin relationship, the average network size was 7.5, and 61% of participants routinely received fictive kin support. Affective closeness and contact with family, friends, and church members were positively associated with fictive kin relations. Age, region, income and marital and parental status were related to fictive kin network involvement, though these associations varied by race/ethnicity. Collectively, findings indicate that fictive kin ties extend beyond marginalized communities, and they operate as a means to strengthen family bonds, rather than substitute for family deficits.

Introduction

Fictive kin play an important role in the familial lives of numerous African Americans (e.g., Aschenbrenner, 1975; Billingsley, 1992; Stack, 1974), Black Caribbeans (Shaw, 2008) and non-Latino Whites (Nelson, 2020). Fictive kin are individuals who, although unrelated by blood or marriage, regard one another as relatives (Sussman, 1976). As such, fictive kin are accorded many of the same rights and statuses as family members and have many of the same duties and responsibilities (Chatters et al., 1994; Stack, 1974). In fact, fictive kin arrangements have been known to function as key sources of social and economic capital among marginalized populations (Muraco, 2006; Ebaugh & Curry, 2000), and some studies indicate that fictive kin can play as central a role in families as persons related by consanguinity (e.g., Dilworth-Anderson, 1992). Despite the acknowledged significance of fictive kin, there remains a paucity of research in this area. Most studies of fictive kin relationships are qualitative in nature and rely on nonrandom samples, while only a handful of studies investigate fictive kin using data from a national probability survey (Chatters, et al., 1994; Taylor et al., 2013). The present study investigates the correlates of fictive kin relationships among a national sample of African Americans, Black Caribbeans, and non-Latino Whites. It extends previous research by investigating within group differences in fictive kin relationships among these three populations. The literature review begins with a discussion of research on fictive kin, followed by research specifically focused on African American, Black Caribbean and non-Latino White fictive kin relationships. Following this, we discuss the theoretical orientation guiding this analysis and the focus of our investigation.

Background and Theoretical Orientation

Defining Fictive Kin

Fictive kinship is described in the literature using a variety of terms (e.g., pseudo-, chosen and voluntary kin, quasi-kin, step-kin) and in relation to different individuals, groups and circumstances such as peers or friends (Alexakos et al., 2011; MacRae, 1992; Rubenstein et al., 1991; Tatum, 1987), extended family (Rogler, 1978; Johnson & Barer, 1990), and church-based networks (Lincoln & Mamiya, 1990). Further, fictive kinship practices have been examined among sexual and gender minorities (Muraco, 2006; Weeks, Heaphy, & Donovan, 2001; Weston, 1991), immigrant communities (Ebaugh & Curry, 2000), and in relation to marital transitions (e.g., step-kin, step-grandparents) such as divorce and remarriage (Allen et al., 2011).

Discussions of kinship extension practices (Braithwaite et al., 2010; Nelson, 2014) note differences in the inherent meanings and uses of terms found in the literature (e.g., chosen-, discretionary-, voluntary-kin), as well in common usage (e.g., ‘play’ relatives, ‘going for brothers’). Differences in terms and meanings are particularly significant when these concepts refer to diverse population groups (i.e., race/ethnic minorities, sexual minorities, White heterosexuals). Chatters et al. (1994) argued that although the variety of recognized terms complicates a broader understanding of the prevalence and functions of kinship extension practices, fictive kin relationships exist across diverse population groups. Underscoring the pervasiveness of fictive kinship practices, Taylor et al., (2013) found that roughly 9 out of 10 adults indicated that they had a fictive kin relation.

African American Fictive Kin Relations

Fictive kin relationships figure prominently in ethnographic research on the informal support networks of African Americans. Research conducted over 40 years ago noted the prominence of fictive kin in African American social support networks (Aschenbrenner, 1975; Stack, 1974). It is surprising, then, that direct examination of social support from fictive kin has received so little attention in the informal support literature. Although research (Chatters et al., 1994; Stack, 1974) confirms the overall significance of fictive kin for African American informal social support networks, two major issues limit this literature. First, the majority of studies are drawn from small, geographically restricted samples using qualitative methods (e.g., in-depth interviews, ethnographies). Additionally, many of the classic (e.g., Stack, 1974) as well as contemporaneous (e.g., Desmond, 2012) investigations often focused on low-income, urban populations. Although these studies provide rich information on fictive kin relations, the lack of sample representativeness limits the generalizability of their findings. Because very little research on fictive kin networks is based on broader, representative surveys of the population, our understanding of the general distribution and the demographic and social correlates of these relationships is extremely limited.

A second issue concerns how fictive kin relations, particularly in regard to African Americans, are characterized in the literature. Specifically, because many studies focus on low-income African Americans (e.g., Desmond, 2012; Stack, 1974), the issue of fictive kin is often viewed from the perspective of material or economic need. In essence, fictive kin ties are regarded as an adaptive strategy that is motivated by financial concerns or in response to a lack of ‘true’ kin (defined biologically or legally) who would usually be obligated and expected to provide material and social resources to vulnerable groups. However, recognized material need and the absence of kin resources may not be the only factors that give rise to fictive kin relationships. Accordingly, it is important to examine within-group diversity in the distribution, sociodemographic correlates, and functions of fictive kin networks.

Black Caribbean Fictive Kin Networks

Research on fictive kinship networks among Black Caribbeans is quite sparse. The available research typically discusses the role of fictive kin within the context of immigration. Ethnographic studies note the importance of fictive kin in providing assistance to recent immigrants, particularly those who do not have biological kin in their new country (Chamberlin, 2017). The type of assistance that recent immigrants receive include transportation, room and board, financial support, and guidance on how to navigate the United States. Shaw’s work (2008) on Haitian immigrants is one of the most in-depth studies of fictive kin relationships among Black Caribbeans. He found that some of the more common examples of fictive kinship relationships were “cousins”, and the use of “mother” or “grandmother” for elderly Haitian women. One innovative feature of his study was the use of a diary format to record the helping behaviors of 14 Haitian immigrants over the course of 4 weeks. Of 566 helping occurrences noted, 42% involved family members and 34% involved fictive kin (Shaw, 2008). Consistent with the transnational nature of Black Caribbean families, family members who received aid were geographically dispersed. In contrast, fictive kin who received assistance resided in their same community.

Fictive Kin Among Non-Latino Whites

Research among non-Latino Whites (MacRae, 1992; Matthews, 1986; Nelson, 2020; Rubenstein et al., 1991; Rubin, 1985; Voorpostel, 2012) confirms the presence and importance of fictive kin relationships, particularly for older adults who have social support deficits (Jordan-Marsh & Harden, 2005). However, as compared to studies involving African American populations, very little research examines fictive kin networks of non-Latino Whites. Nonetheless, there are several notable studies of fictive kin relationships focused on this population. A study based on a non-probability sample of older Whites in Nova Scotia found that 40% of respondents reported that they had a fictive kin relation (MacRae, 1992). A recent ethnographic study of 75 middle-class Whites (Nelson, 2020) provides one of the most comprehensive investigations of fictive kin relationships. One of the unique aspects of Nelson’s book is that it solely focused on fictive kinship relations, whereas for most other major books in this area, fictive kin is only one of numerous issues examined (e.g., Aschenbrenner, 1975; Billingsley, 1992; Stack, 1974). Nelson’s study documents the diverse ways that fictive kin relationships are established, contested, and maintained among this population. Further, it explicitly acknowledges that fictive kin relationships are not enacted to merely compensate for perceived deficiencies in family networks. Rather, some middle-class Whites reported that fictive kin ties operated as supplements to, rather than substitutes for, other family relationships.

A study using data from the National Survey of American Life (NSAL) found that non-Latino Whites were less likely than African Americans and Black Caribbeans to report both having fictive kin and to have large fictive kin networks (Taylor et al., 2013). Non-Latino Whites, however, reported receiving support from fictive kin more frequently than either African Americans or Black Caribbeans. The findings suggest that although non-Latino Whites are less likely to have fictive kin in their networks, those with fictive kin may have stronger ties to them. Collectively, these findings indicate that fictive kin are a common feature of African American, Black Caribbean and non-Hispanic White family networks and their role as sources of support is more complex than previously thought (Taylor et al., 2013). To our knowledge, the present study is the first investigation of within-group differences in fictive kin network among non-Latino Whites based upon a probability sample.

In summary, fictive kin relationships and networks are found among diverse groups of the population, including African Americans, Latinos, non-Latino Whites, older adults, and sexual minority families (Chatters et al., 1994; Jordan-Marsh & Harden, 2005; Nelson, 2020; Taylor et al., 2013). Across race, cultural, and identity groups, fictive kin relationships are accorded specific family statuses as well as associated role behaviors and expectations (Chatters et al., 1994; Gill-Hopple & Brage-Hudson, 2012; Stewart, 2008). Information about the circumstances surrounding kinship extension and the consequences and benefits deriving from fictive kin are important for understanding how families respond and adapt to the changing circumstances, expectations, and needs of family members (Allen et al., 2011; Nelson, 2020; Reid & Reczek, 2011; Voorpostel, 2012). Accordingly, research on the prevalence and correlates of fictive kin relations can help us better understand the personal, social, and cultural context of fictive kin relationships within diverse groups.

The Family Solidarity Model

We use the family solidarity model as the theoretical orientation for our analysis of fictive kin relationships. This well-established model of family support, first articulated and tested by Bengtson and colleagues (1988, 1991, 1995, 1999), views the emotional bonds between individuals as a fundamental organizing feature of family life. This model identifies a set of dimensions that characterize family relations, such as frequency of interaction, feelings of closeness, and receipt of support (Bengtson, Giarrusso, Mabry, & Silverstein, 2002; McChesney & Bengtson, 1988; Nye & Rushing, 1969). Importantly, the family solidarity model proposes that key features of family relationships like family contact and family closeness are positively correlated with another and are also positively related to exchanges of support. The use of the family solidarity model may seem counterintuitive for an analysis of non-kin. However, mechanisms that facilitate family support networks also facilitate non-kin networks. That is, friends who are subjectively closer and interact more frequently will be more likely to receive support from their friendship network. Research using this theoretical model has also been applied to non-kin networks. For example, research on church support networks of African Americans finds that, similar to family support, interaction with church members and degree of subjective closeness to church members is associated with receiving more frequent support from church members (Taylor, Lincoln, & Chatters, 2005).

The Focus of the Present Investigation

The current study investigates the social and demographic correlates of fictive kin involvement among African Americans, Black Caribbeans, and non-Latino Whites. It is based on the National Survey of American Life, a nationally representative sample of African Americans, Black Caribbeans, and non-Latino Whites. Specifically, we examine three questions related to fictive kin involvement: what factors shape 1) whether or not one has fictive kin, 2) the number of fictive present kin in one’s network, and 3) the frequency of support received from fictive kin? The social and demographic correlates of fictive kin networks examined in this analysis include frequency of contact with family, friends, and church members, feelings of closeness to family and friends, commonly used demographic characteristics (e.g., age, gender, and education), as well as expanded information on marital and partner status (i.e., remarriage, cohabitation).

It bears emphasizing that there are many excellent conceptual articles (e.g., Nelson, 2014), literature reviews (e.g., Jordan-Marsh & Harden, 2005; Nelson, 2013; Voorpostel, 2012), ethnographic studies (e.g., Aschenbrenner, 1975; Stack, 1974), and other works based on semi-structured or in-depth interviews (e.g., Johnson, 1999; Nelson, 2020). However, the vast majority of research on fictive kin is based on geographically localized convenience samples. This includes investigations involving 14 African American college students (Brooks & Allen, 2016), 45 older adults (Allen et al., 2011), 110 adults (Braithwaite et al., 2010), 71 displaced Hurricane Katrina survivors (Reid & Reczek, 2011), a single large extended family (Stewart, 2008), 23 friendship dyads (Muraco, 2006), 114 caregivers and care recipients (Barker, 2002), 122 African Americans 85 years and older (Johnson, 1999), 75 middle class White adults (Nelson, 2020), and 142 older Canadian women (MacRae, 1992). Additionally, because much of the qualitative research on fictive kin relations among Black Americans has focused on low income communities (e.g., Desmond, 2012; Stack, 1974), it is often assumed that fictive kin are only present among poor African Americans. Consequently, several important questions in this area remain, including whether differences in the prevalence of fictive kin and receipt of fictive kin support differ by socio-economic status, gender, region, or other key demographic or social background factors. To our knowledge, there are only a few studies based on survey data (e.g., Chatters et al., 1994) and, only two that focus on correlates of fictive kin networks (using fictive kin networks as dependent variables) and use population-based national probability samples (Chatters et al., 1994; Taylor et al., 2013).

Our study builds upon these two earlier efforts in two ways. First, we extend Chatters et al.’s (1994) work by investigating not only factors associated with the presence of fictive kin, but also the correlates of the receipt of assistance from fictive kin. The present analysis also examines a larger and more diverse set of independent variables than previously included. Second, this study builds upon Taylor et al.’s (2013) investigation of racial and ethnic differences in extended family, friend, congregational, and fictive kin networks by focusing on within-group heterogeneity in fictive kin relations. Together, our analyses expand upon previous efforts by providing the first comprehensive within-group analysis of fictive kin relationships among African Americans, Black Caribbeans and non-Latino Whites, in order to illuminate the diversity that exists in fictive kin relationships within these populations.

Study Expectations

Although to our knowledge, no previous studies systematically examine within-group variation in fictive kin networks among these populations, the literature provides useful information about how these relationships are governed. Based on this information, we propose several expectations concerning the factors that shape the quantity of fictive kin present in individuals’ networks and the frequency of support that they receive. First, fictive kinship is defined as a process involving the intensification of friendship bonds and expectations (Chatters et al., 1994). Accordingly, we anticipate that higher levels of subjective closeness and contact with friends will be positively associated with having fictive kin, the number of fictive kin identified, and levels of social support. Social relationships between church congregants are often discussed in fictive kinship terms. For example, research on African Americans has found that members of one’s church are often called ‘brother’ or ‘sister’, congregants refer to their church members as their ‘church family,’ and designations such as ‘church mother’ are often used as honorific titles for persons who are deemed to be important and respected church members (Lincoln & Mamiya, 1990; Taylor, Levin & Chatters, 2004). Consequently, we also expect that individuals who report more contact with church members will be more likely to have fictive kin, report higher levels of fictive kin relations, and receive more fictive kin support.

With respect to relationships between family closeness and contact and fictive kin, two alternatives are possible. If, as suggested in the literature, fictive kinship involves a general intensification of family bonds, we would expect that family closeness and contact are positively associated with having fictive kin and received support from fictive kin. This would be consistent with the family solidarity model. Alternatively, fictive kinship practices are often characterized as a strategy to compensate for deficits in the kin system that requires the substitution of non-kin (Cantor, 1979). If this is the case, we anticipate that persons who are not emotionally close to their family and have little contact with them will report higher levels of fictive kin involvement (i.e., having fictive kin, the number identified, and support) than their counterparts (manifested in negative associations between family and fictive kin factors).

As it pertains to relationships between demographic background characteristics and fictive kin relations, there are two possibilities. A need-based argument for the incorporation of fictive kin would suggest that individuals who have lower incomes, lower levels of education, are parents, or are older would be more likely to be involved with fictive kin. Conversely, if fictive kin relations largely reflect cultural preferences or are a more normative experience, then we would not expect to observe statistically significant differences in fictive kin networks by access to economic resources, age, or parental status.

With regards to marital status differences, one line of research suggests that marital transitions (e.g., separation, divorce) that result in changes in recognized kin relationships may require an informal redefinition of familial associations (Allen et al., 2011). For example, former in-laws who remain affectively close and a member of one’s social network may be redefined as fictive kin. Accordingly, individuals who experience divorce or separation may report a greater likelihood of having fictive kin, larger numbers of fictive kin, and receiving support than their married counterparts. On the other hand, married persons have access to their spouse’s kin and friendship networks as well as their own. As a result, they may have a larger pool of individuals who could be defined as fictive, which could result larger fictive kin networks.

Data and Methods

Data

The National Survey of American Life: Coping with Stress in the 21st Century (NSAL) was conducted by the Program for Research on Black Americans at the University of Michigan’s Institute for Social Research. The field work for the study was completed by the Institute for Social Research’s Survey Research Center, in cooperation with the Program for Research on Black Americans. The NSAL sample has a national multi-stage probability design which consists of 64 primary sampling units (PSUs). Fifty-six of these primary areas overlap substantially with existing Survey Research Center’s National Sample primary areas. The remaining eight primary areas were chosen from the South in order for the sample to represent African Americans in the proportion in which they are distributed nationally. The African American sample is a national representative sample of households located in the 48 coterminous states and Washington D.C. with at least one Black adult aged 18 years or older who did not identify ancestral ties in the Caribbean. Both the African American and non-Latino White samples were selected exclusively from these targeted geographic segments in proportion to the African American population (Herringa et al., 2004).

The NSAL includes the first major probability sample of Black Caribbeans. For the purposes of this study, Black Caribbeans are defined as persons who trace their ethnic heritage to a Caribbean country, but who now reside in the United States, are racially classified as Black, and who are English-speaking (but may also speak another language). In both the African American and Black Caribbean samples, it was necessary for respondents to self-identify their race as Black. Those self-identifying as Black were included in the Black Caribbean sample if they: a) answered affirmatively when asked if they were of West Indian or Caribbean descent, b) said they were from a country included on a list of Caribbean area countries presented by the interviewers, or c) indicated that their parents or grandparents were born in a Caribbean area country.

The data collection was conducted from February 2001 to June 2003. Most of the interviews were conducted face-to-face (86%) in respondents’ homes, whereas the remaining 14% were telephone interviews; respondents were compensated for their time. A total of 6,082 face-to-face interviews were conducted with persons aged 18 or older. The overall response rate was 72.3%. Final response rates for the NSAL two-phase sample designs were computed using the American Association of Public Opinion Research (AAPOR) guidelines (for Response Rate 3 samples) (AAPOR 2006) (see Jackson et al. 2004 and Herringa et al., 2004 for a more detailed discussion of the NSAL sample). The NSAL data collection was approved by the University of Michigan Institutional Review Board.

Dependent Variables

Three dependent variables measure involvement in fictive kin networks: having fictive kin, number of fictive kin, and frequency of receiving support from fictive kin. Number of fictive kin is assessed by the question: “How many people are close to your family who are not really blood-related or marriage-related but who are treated just like a relative?” Using this question, the dichotomous variable “having a fictive kin” was created, which contrasts respondents who indicate having at least one fictive kin compared to those who report none. Frequency of receiving support from fictive kin is measured by the item: “How often do they (fictive kin) -- help you out?” Values for response categories are very often=4, fairly often=3, not too often=2 and never=1.

Independent Variables

Five variables are used to measure involvement in extended family, friendship, and church informal social support networks. Two measures assess involvement in family support networks. Degree of subjective family closeness is measured by the question: “How close do you feel towards your family members?” Values for response categories are very close=4, fairly close=3, not too close=2 and not close at all=1. Frequency of contact with family members is measured by the question: “How often do you see, write or talk on the telephone with family or relatives who do not live with you?” Values for response categories are nearly everyday=7, at least once a week=6, a few times a month=5, at least once a month=4, a few times a year=3, hardly ever=2 and never=1. Similarly, closeness to friends (i.e., subjective friendship closeness) and frequency of contact with friends (i.e., friend contact) are assessed using the same response format as the family closeness and contact questions. Frequency of contact with church members is measured by the question: “How often do you see, write or talk on the telephone with members of your church (place of worship)? Respondents who interacted with church members nearly everyday received a 7, the highest score on this variable. This was followed by persons who report interacting with their church members at least once a week (6), a few times a month (5), and a few times a year (4). Persons who never interact with their church members received a score of 3 and respondents who attend religious services less than once a year received a score of 2. Finally, persons who have not attended religious services (excluding weddings and funerals) since the age of 18 received a score of 1.

Sociodemographic variables (i.e., age, gender, family income, education, marital status, region, and parental status) are utilized as independent variables. Age and education are coded in years and family income is coded in dollars. The staff of the Program for Research on Black Americans imputed missing data for education for 74 cases (1.2% of the total NSAL sample) and for income for 773 cases (12.7% of the total NSAL sample). In the multivariate analysis, income has been divided by 5000 in order to provide a better understanding of the net impact of income. Marital status is coded as married (reference), remarried, cohabiting, divorced, widowed, separated, and never married. Parental status is coded as parent (reference) or non-parent. Region is coded as South (reference), Northeast, North Central, and West.

Additionally, two variables that are particularly relevant to Black Caribbeans in the United States are included in the analysis of this population: country of origin and immigration history. Respondents identified over 25 different countries of origin. Country of origin was recoded into five categories: Jamaica, Trinidad-Tobago, Other English-speaking country (e.g., Barbados), Spanish-speaking country (e.g., Puerto Rico, Dominican Republic), and Haiti. Immigration status has five categories corresponding to respondents who were: 1) born in the United States (reference), 2) immigrated to the United States 0 to 5 years ago, 3) immigrated to the United States 6–10 years ago, 4) immigrated to the United States 11–20 years ago, and 5) immigrated to the United States more than 20 years ago.

Analysis Strategy

All analyses were conducted with SAS 9.13, which uses the Taylor expansion approximation technique for calculating the complex-design based estimates of variance; in addition, all tests of significance are complex sample design-corrected estimates. Logistic regression was used with the dichotomous dependent variable of whether a respondent has fictive kin. Ordinary least squares regression analysis was used with the two continuous dependent variables (number of fictive kin and receipt of support from fictive kin). Because the number of fictive kin variable was not normally distributed, the log of this variable was used. After the logarithmic transformation, this variable has a normal distribution. Data used in these analyses are weighted to correct for unequal probabilities of selection and non-response. Lastly, we computed the Variance Inflation Factor (VIF) to check for multicollinearity among the independent variables for all of the analyses. The largest VIF was 2.77, which is far below both the threshold of 10 and the more stringent threshold of 4, which many researchers regard as a sign of severe or serious multicollinearity (O’Brien, 2007).

Results

Descriptive Statistics

Table 1 presents the demographic description of the sample and the distribution of the study variables. The sample is almost evenly split between males (46%) and females (54%) and the average age is 44 years old. The mean household income is roughly $42,000 and the average years of education is 13. Just over half the sample is married (28%) or never married (27%), and a quarter are remarried (13%), or divorced (12%). With respect to our key outcome variables, we observe that 87% of respondents report that they have at least one fictive kin. The average number of fictive kin is 7.5. Among respondents who have fictive kin, 24% indicate that they receive support from them very often, 37% report fairly often, 27% report not too often, and 11% report never receiving support.

Table 1.

Demographic characteristics of the sample and distribution of study variables, National Survey of American Life (NSAL)

Variables	Percent	N	Mean	S.D.	Min	Max
Has fictive kin	87.43	5260
No. of Fictive Kin		5901	7.54	12.03	0	100
0	12.56	635
1	6.79	441
2	11.19	722
3	10.77	617
4	8.70	531
5	12.14	718
6	5.73	348
7	1.83	128
8	3.09	142
9	0.29	28
10	11.71	672
11–15	5.98	287
16–20	4.18	276
21–100	5.02	356
Receipt of support from fictive kin
Never	11.07	612
Not too often	27.35	1482
Fairly often	37.34	1716
Very often	24.24	1116
Female	54.13	3796
Age		6082	43.57	16.61	18	94
Family income		6082	42417.66	39411.54	0	650000
Education		6082	12.89	2.65	0	17
Marital status
Married	27.52	1391
Remarried	12.73	511
Cohabiting	7.81	435
Separated	5.08	451
Divorced	12.31	849
Widowed	7.74	534
Never married	26.81	1885
Region
Northeast	20.56	1653
Midwest	12.91	690
South	54.48	3395
West	12.06	344
Immigration status*
Born in the U.S.	35.76	440
Immigrated 0–5 years ago	7.89	119
Immigrated 6–10 years ago	8.25	164
Immigrated 11–20 years ago	19.91	364
Immigrated >20 years ago	28.19	512
Country of origin*
Jamaica	9.69	510
Spanish-speaking country	5.92	180
Haiti	3.78	298
Trinidad-Tobago	3.91	170
Other	8.29	440
Is a parent	74.89	4874
Church member contact		6075	4.15	1.89	1	7
Family contact		6029	6.05	1.22	1	7
Family closeness		6026	3.63	0.64	1	4
Friendship contact		5954	5.79	1.5	1	7
Friendship closeness		5949	3.32	0.74	1	4

Notes: Percents and sample size are presented for categorical variables and means and standard deviations are presented for continuous variables. Percentages are weighted and frequencies are unweighted. N=number of observations. S.D.=standard deviation. Min.=minimum. Max=maximum.

^*=Information available for Black Caribbean sample only.

Multivariate Results

African Americans.

The multivariate analysis of the family, friendship, church member, and demographic correlates of fictive kin networks among African Americans are presented in Table 2. With respect to having fictive kin (Model 1), age, income, church member contact, subjective closeness to family, and subjective closeness to friends are significant correlates. Older respondents are less likely than younger persons to have fictive kin, while persons with higher incomes are more likely to report having fictive kin. Respondents who have more frequent contact with church members and those who are subjectively closer to their family and to their friends are also substantially more likely to have fictive kin.

Table 2.

Logistic and ordinary least squares regression analysis identifying correlates of fictive kin network involvement among African Americans, National Survey of American Life (NSAL)

	Model 1		Model 2		Model 3
Variables	Has fictive kin		No. of fictive kin		Receipt of support from fictive kin
	β	SE	β	SE	β	SE
Age	−0.02***	0.01	−0.01**	0.00	−0.01***	0.00
Female (vs. male)	0.04	0.08	−0.07	0.04	0.02	0.04
Family income	0.05*	0.02	0.01*	0.00	0.00	0.00
Years of education	0.05	0.03	0.00	0.01	−0.05***	0.01
Region (vs. South)
Northeast	0.04	0.16	0.09	0.05	0.05	0.07
North Central	0.04	0.18	0.06	0.06	0.04	0.06
West	−0.13	0.17	−0.04	0.07	−0.23***	0.05
Marital status (vs. married)
Remarried	−0.28	0.21	−0.10	0.06	−0.01	0.08
Cohabiting	0.33	0.26	−0.02	0.07	−0.04	0.09
Separated	−0.20	0.26	−0.13	0.10	0.14	0.08
Divorced	−0.09	0.18	−0.05	0.05	−0.11	0.06
Widowed	−0.01	0.23	−0.11	0.07	−0.04	0.1
Never Married	−0.11	0.17	−0.06	0.07	0.00	0.04
Is a parent	0.09	0.11	0.14*	0.05	−0.05	0.05
Church member contact	0.15***	0.04	0.08***	0.01	0.04**	0.01
Family contact	0.07	0.06	0.04*	0.02	−0.04	0.02
Family closeness	0.36**	0.11	0.14***	0.03	0.06	0.04
Friendship contact	0.10	0.05	0.02	0.01	0.07***	0.01
Friendship closeness	0.43***	0.11	0.19***	0.03	0.34***	0.17
Constant	−1.77***	0.53	0.03	0.20	1.89
F			44.83***		70.36***
R2	0.05		0.11		0.14
N	3366		3366		2879

^*p < .05

^**p< .01

^***p < .001

Notes: Data are weighted to account for unequal probabilities of selection, non-response, and post-stratification. Logistic regression was employed in the analysis predicting whether respondents have fictive kin. Ordinary least squares regression analysis was used in the analysis identifying correlates of number of fictive kin and receipt of support from fictive kin. No.=number.

Analysis of the number of fictive kin (Table 2, Model 2) reveals that age, income, parental status, frequency of contact with church members, frequency of contact with family members, subjective closeness to family, and subjective closeness to friends are significant correlates. Older respondents report having fewer fictive kin than younger respondents. African Americans with higher incomes are more likely to report having fictive kin than those with lower incomes. Parents reported having more fictive kin in their networks than their counterparts. Church member contact and family contact, as well as subjective family and friendship closeness are all positively associated with number of fictive kin.

The analysis for frequency of receiving support from fictive kin is presented in Table 2 (Model 3). Age, education, region, church member contact, friendship contact, and closeness to friends are significantly associated with frequency of receiving support from fictive kin. Older respondents report that they receive assistance from fictive kin less frequently than younger persons. Respondents with more years of education report receiving assistance less frequently than their counterparts. Respondents residing in the West receive support from fictive kin less frequently than Southerners. Frequency of contact with church members, frequency of contact with friends, and friendship closeness are all positively associated with frequency of fictive kin support.

Black Caribbeans.

The regression analysis of fictive kin for Black Caribbeans is presented in Table 3 Model 1. Divorced respondents are significantly less likely to have fictive kin than married respondents. Church member contact, subjective family closeness, and friendship closeness are positively associated with having fictive kin. Gender, marital status, country of origin, church member contact, and family closeness are significantly associated with number of fictive kin (Table 3, Model 2). Women and remarried respondents have smaller fictive kin networks than their male and married counterparts. Haitians have smaller fictive kin networks than Jamaicans. Church member contact and family closeness are positively associated with the number of fictive kin. With regards to the receipt of support (Table 3, Model 3) divorced respondents received support less frequently than their married counterparts. In addition, frequency of contact with friends and subjective friendship closeness are positively associated with frequency of receiving fictive kin support.

Table 3.

Logistic and ordinary least squares regression analysis identifying correlates of fictive kin network involvement among Black Caribbeans, National Survey of American Life (NSAL)

	Model 1		Model 2		Model 3
Variables	Has fictive Kin		No. of fictive kin		Receipt of support from fictive kin
	β	SE	β	SE	β	SE
Age	−0.01	0.01	−0.01	0.00	−0.01	0.00
Female (vs. male)	−0.18	0.16	−0.28**	0.09	0.16	0.09
Family income	0.03	0.02	0.01	0.01	0.00	0.00
Years of education	−0.05	0.07	−0.01	0.01	0.00	0.02
Marital status (vs. married)
Remarried	−0.29	0.36	−0.19*	0.09	0.23	0.17
Cohabiting	0.88	0.63	−0.11	0.19	−0.18	0.20
Separated	−0.28	0.50	−0.17	0.16	−0.24	0.15
Divorced	−0.98*	0.47	−0.36	0.24	−0.38*	0.18
Widowed	−0.52	0.64	−0.08	0.25	0.25	0.30
Never Married	0.59	0.33	−0.08	0.16	−0.21	0.18
Is a parent	0.19	0.23	0.15	0.14	0.00	0.13
Immigration status (vs. U.S. born)
Immigrated 0–5 years ago	−0.19	0.49	0.20	0.39	0.03	0.18
Immigrated 6–10 years ago	0.05	0.32	−0.10	0.15	−0.10	0.09
Immigrated 11–20 years ago	0.15	0.24	−0.01	0.10	−0.06	0.08
Immigrated >20 years ago	−0.04	0.30	−0.08	0.19	−0.09	0.10
Country of origin (vs. Jamaica)
Spanish-speaking country	−0.14	0.46	0.32	0.21	−0.03	0.13
Haiti	−0.49	0.29	−0.19*	0.08	−0.08	0.11
Trinidad-Tobago	0.57	0.29	0.26	0.24	−0.08	0.12
Other	−0.01	0.28	0.38***	0.1	−0.04	0.09
Church member contact	0.21*	0.09	0.09***	0.02	−0.01	0.02
Family contact	0.04	0.08	0.00	0.03	0.02	0.05
Family closeness	0.73***	0.18	0.32***	0.05	0.11	0.06
Friendship contact	0.14	0.10	0.04	0.04	0.13***	0.04
Friendship closeness	0.62*	0.22	0.08	0.10	0.26***	0.06
Constant	−3.40**	1.20	0.08	0.30	1.10*	0.45
F			437.61***		489.09***
R2	0.09		0.19		0.15
N	1498		1503		1253

^*p < .05

^**p< .01

^***p < .001

Notes: Data are weighted to account for unequal probabilities of selection, non-response, and post-stratification. Logistic regression was employed in the analysis predicting whether respondents have fictive kin. Ordinary least squares regression analysis was used in the analysis identifying correlates of number of fictive kin and receipt of support from fictive kin. No.=number.

Non-Latino Whites.

Table 4 presents the analysis of fictive kin for non-Latino Whites. The regression coefficients in Model 1 indicate that marital status, parental status, subjective family closeness, friendship contact, and friendship closeness are significantly associated with having fictive kin. Separated respondents are less likely than marrieds to have fictive kin, whereas parents are more likely than their counterparts to have fictive kin. Subjective family closeness, friendship contact, and friendship closeness are all positively associated with having fictive kin. Model 2 (Table 4) presents analysis for number of fictive kin. Marital status, parental status, church member contact, subjective family closeness, and friendship closeness are significantly associated with number of fictive kin. Divorced respondents have fewer fictive kin, whereas parents have more fictive kin than their respective counterparts. Respondents who have more frequent contact with their church members, greater levels of subjective family closeness, and greater levels of friendship closeness have more fictive kin. Marital status, church member contact, subjective family closeness, and friendship closeness are significantly associated with the receipt of support from fictive kin (Model 3). Respondents in cohabiting relationships are more likely to receive support from fictive kin than married respondents. Church member contact, subjective family closeness, and friendship closeness all have significant positive relationships with frequency of receiving fictive kin support among non-Latino Whites.

Table 4.

Logistic and ordinary least squares regression analysis identifying correlates of fictive kin network involvement among White Americans, National Survey of American Life (NSAL)

	Model 1		Model 2		Model 3
Variables	Has fictive kin		No. of fictive kin		Receipt of support from fictive kin
	β	SE	β	SE	β	SE
Age	−0.02	0.01	0.00	0.00	0.00	0.00
Female (vs. male)	−0.14	0.16	−0.14	0.08	0.01	0.11
Family income	0.02	0.02	0.00	0.00	−0.01	0.01
Years of education	−0.05	0.04	0.00	0.01	−0.02	0.02
Region (vs. South)
Northeast	−0.06	0.31	−0.07	0.12	0.14	0.07
North Central	0.21	0.32	0.00	0.07	0.05	0.14
West	−0.08	0.31	0.12	0.06	−0.16	0.15
Marital sttatus (vs. married)
Remarried	−0.09	0.49	−0.16	0.12	−0.13	0.17
Cohabiting	0.59	0.5	0.02	0.16	0.67***	0.12
Separated	−0.80*	0.37	−0.36	0.21	−0.05	0.27
Divorced	−0.41	0.3	−0.33**	0.10	−0.19	0.12
Widowed	−0.28	0.39	−0.20	0.20	−0.18	0.17
Never Married	0.72	0.45	0.03	0.12	0.27	0.16
Is a parent	0.41*	0.19	0.25**	0.07	0.23	0.14
Church member contact	0.09	0.07	0.07**	0.02	0.06*	0.02
Family contact	0.07	0.1	0.03	0.03	−0.04	0.04
Family closeness	0.37*	0.18	0.20*	0.07	0.17**	0.05
Friendship contact	0.17*	0.07	0.04	0.02	0.05	0.04
Friendship closeness	0.71***	0.19	0.34***	0.07	0.29**	0.09
Constant	−2.47*	1.08	−0.88	0.36	1.14*	0.39
F			52.08***		181.00**
R2	0.37		0.21		0.17
N	848		849		677

^*p < .05

^**p< .01

^***p < .001

Notes: Data are weighted to account for unequal probabilities of selection, non-response, and post-stratification. Logistic regression was employed in the analysis predicting whether respondents have fictive kin. Ordinary least squares regression analysis was used in the analysis identifying correlates of number of fictive kin and receipt of support from fictive kin. No.=number.

Discussion

Descriptively, the vast majority of African Americans, Black Caribbeans and non-Latino Whites reported having at least one fictive kin (87%) and, on average, had 7.54 people who were regarded as fictive kin. This estimate for number of respondents with fictive kin is higher than that reported by Chatters et al., (1994) who found that 66% of African American respondents in the National Survey of American Life indicated having fictive kin. More than half of persons with fictive kin indicated that they received social support on a frequent basis (combining the categories very and fairly often), 27% indicated receiving assistance not too often, and 11% reported that they never received support from fictive kin. The remainder of this section first addresses the social integration correlates of fictive kin, followed by a discussion of the demographic correlates.

Overall, our findings verified the importance of family, friend, and congregational involvement for fictive kin relations for all three populations. In general, higher levels of affective closeness and contact with family, friends, and church members were positively associated with the presence of fictive kin, having a greater number in one’s network, and receiving more frequent assistance from them. These positive relationships are consistent with the family solidarity model, which notes that positive family relations facilitate family contact and receipt of family support (Bengtson, Giarrusso, Mabry, & Silverstein, 2002). Taken as a whole, these positive associations support the notion of kinship extension as a means to strengthen family bonds, rather than as a strategy to substitute for family relations based on consanguinity or marriage (Nelson, 2020).

Study findings revealed that the relationships between family, friendship, and church member involvement and having a fictive kin, number of fictive kin, and receiving fictive support were mostly consistent across all three populations. Because there are 3 dependent variables and 3 different populations, there are 9 (3×3) potential significant relationships between, for example, church member contact and the fictive kin outcome variables. Church member contact was significant in 7 of the 9 regressions, suggesting that church networks are an important source of fictive kin relationships. This is likely due to attitudes and practices that regard fellow congregants as part of one’s church family (Johnson & Barer, 1990; Lincoln & Mamiya, 1990; Taylor et al., 2013). Persons in frequent contact with church members have more access to church-based networks that promote the development of fictive kin and, as a consequence, result in a greater number of these relationships (Taylor et al., 2005). For instance, in many both African American and non-Latino White Christian churches, it is not unusual for individuals to greet each other using family pronouns such as ‘brother’, ‘sister’, or ‘mother’. Similarly, in African American communities, it is not unusual for local residents to ask individuals who have recently moved to that town if they have found a ‘church family’. This is commonly understood as another way of asking whether they have found a church that they like.

Significant findings for the family and friendship variables revealed that subjective closeness was more likely to be significantly associated with fictive kin than frequency of contact. Family closeness was significant in 7 of the 9 regressions, whereas family contact was significant in only 1 of the 9 regressions. Similarly, friendship contact was significant in 3 of the 9 regressions and friendship closeness was significant in 8 of the 9 regressions. This analysis shows that the affective dimensions of family and friendship relationships are more important than mere contact. The role of family closeness and friend closeness for fictive kin may reflect a general perception of connection and rapport with members of these groups that, in turn, facilitates the development of fictive kin ties.

Although there are some variations in the family, friend, and church member correlates of the receipt of support from fictive kin by race/ethnicity, these differences are relatively minor. For Black Caribbeans, relationships with friends (but not family or church members) are associated with receipt of fictive kin support. For African Americans, relationships with friends and church members are related to receiving support. Thus, for African Americans and Black Caribbeans relationships with non-kin (friends and church members) were more important than family ties with regards to receiving support from fictive kin. This was not the case among non-Latino Whites where relationships with church members, friends, and family were associated with receiving support from fictive kin. Overall, the comparative findings indicate that despite some differences in the correlates of fictive kin relations, African Americans, Black Caribbeans and non-Latino Whites who are more integrated in their family, friendship, and church member networks have a higher likelihood of also being integrated in their fictive kin networks.

Finally, some studies suggest that deficits in family resources motivate seeking assistance from fictive kin (Anderson, 1978; Desmond, 2012). This is particularly the case for research on poor and financially struggling African Americans (e.g., Anderson’s street corner men, Desmond’s evicted urban poor). If that were the case, in our analysis, low levels of family closeness and contact (as proxies for family resources) would be associated with more frequent support from fictive kin (i.e., a negative relationship). Our findings, however, indicated that family contact and closeness were, in fact, either irrelevant or in the one instance where it was significant, positively associated with receiving fictive kin support; it was only among non-Latino Whites that family closeness was significantly associated with receiving fictive kin support.

The analyses indicated several notable findings related to demographic background characteristics. Among African Americans, age was negatively associated with the likelihood of having fictive kin, number of fictive kin, and frequency of receiving support from fictive kin. These negative age findings are consistent with previous research on African American fictive kin networks indicating that older adults have less fictive kin involvement (Chatters et al., 1994). Among Black Caribbeans and non-Latino Whites, however, age was not significantly associated with any of the fictive kin variables. Combined, these findings are consistent with recent research on the impact of racial disparities in mortality on family and friendship relationships (Umberson, 2017). That is, because African Americans have lower life expectancies, they experience more losses of family and friends over the life course in comparison to non-Latino Whites (Umberson, 2017). In the present analysis, the lower levels of involvement in fictive kin networks among older African Americans is likely due to the deaths of fictive kin.

As stated previously, much of the classic (e.g., Stack, 1974) as well as contemporaneous (e.g., Desmond, 2012) research on fictive kin among African Americans was conducted in low-income communities. Given the focus on these studies, a perception emerged that fictive kin networks were more prevalent among low-income African Americans. However, our analysis, in conjunction with previous research on African Americans (Chatter et al., 1994), indicates that this is not the case. Our findings based on a national probability sample indicate that fictive kin ties occur across all income levels and do not exclusively occur in low-income populations. As a matter of fact, we found that among African Americans, income was positively associated with the likelihood of both having fictive kin and the number of fictive kin. Further, income was not associated with any of the fictive kin variables among Black Caribbeans and non-Latino Whites. In the present study, among African Americans only, lower levels of education were associated with receiving support more frequently from fictive kin. This is the only finding in our analysis that is consistent with the assumption that African Americans with lower socio-economic status are more involved with fictive kin. Collectively, our findings indicate that for African Americans, higher incomes are associated with both having fictive kin and a larger fictive kin network, whereas fewer years of formal education is associated with more frequent support transfers from fictive kin.

One notable finding that emerged from the present study is that African Americans who were parents had more fictive kin than their childless counterparts. Similarly, among non-Latino Whites, parents were more likely to have fictive kin and had more fictive kin than their counterparts. These findings are concordant with research indicating that children expand the informal networks of their parents. This is especially evident for parents who have adult children who fulfill an important role in maintaining and expanding their parents’ networks (Chatters et al., 1985). Parents may feel that it is inappropriate for children to address their friends using their first names and they may find the terms Mr. or Ms. too formal. Accordingly, they instruct their children to refer to very close adult family friends by kinship terms such as ‘aunt’ or ‘uncle’. In addition, Chatters et al. (1994) notes that among African Americans, godparents are generally referred to as ‘aunt’ or ‘uncle’.

There were a few marital status differences in fictive kin relations among Black Caribbeans and non-Latino Whites, but none were found for African Americans. Among Black Caribbeans, divorced respondents were less likely to have fictive kin and received assistance from fictive kin less frequently. Black Caribbeans who were remarried had smaller fictive kin networks compared to their married counterparts. Among non-Latino Whites, those who were separated were less likely to have fictive kin, divorced respondents had fewer fictive kin, and those who were cohabiting received assistance from fictive kin more frequently than their married counterparts. Collectively, these findings are inconsistent with the perspective that unmarried respondents are more involved in fictive kin networks. Conversely, our findings are partially consistent with the view that married persons have access to their own and spouse’s kin and friendship networks and, as a result, are advantaged in forming fictive kin relations. It is important, however, to note that overall, there were very few marital status differences. In addition, our findings do not align with Allen et al.’s (2011) work on the incorporation of divorced in-laws in the extended network. However, our analysis is not suited to detect the nuanced redefinitions by which ex-in-laws come to be regarded as fictive kin that Allen et al., (2011) found in their qualitative research.

Finally, among African Americans, those in the South received assistance from their fictive kin more frequently than those residing in the West. This latter finding of more frequent assistance is consonant with previous research among African Americans indicating that Southerners have several support advantages and receive support more frequently from extended family (Taylor, 1985), have a larger network of kin and non-kin who would help them when ill (Chatters et al., 1985) and have family network typologies that are characterized as optimal as opposed to ambivalent (Taylor et al., 2014).

Several limitations of this study should be considered. The NSAL data is cross-sectional, and, as such, cannot identify causal relationships or attend to issues related to bidirectional relationships. The availability of prospective information would shed light on how demographic, family, church member and friendship characteristics are causally linked in the development and functioning of fictive kin relationships. The Black Caribbean sample excludes individuals who do not speak English (i.e., persons who only speak Haitian-French dialect) and, as a consequence, the study findings are not generalizable to these subgroups. Additionally, the non-Latino White sample does not reflect the regional distribution of the White population, and instead reflects the regional distribution of the African American population. Future investigations should seek to identify the types of assistance that are most commonly received from fictive kin and the correlates of the receipt of support. This information would be helpful in establishing which types of instrumental and emotional support are routinely received versus those that are deployed in response to a crisis situation. Notwithstanding these limitations, the present study is the most comprehensive quantitative investigation to date on the prevalence and correlates of fictive kin.

In conclusion, findings from this study and others using the NSAL show that fictive kin networks are quite pervasive: approximately nine out of ten African Americans and Black Caribbeans as well as eight out of ten non-Latino Whites (Taylor et al., 2013) have at least one fictive kin relationship. These studies confirm that fictive kin relationships are not limited to low income communities, but rather are normative among African Americans, Black Caribbeans and non-Latino Whites in the United States from various socioeconomic backgrounds. Our findings further clarify that fictive kin relationships are associated with a variety of circumstances in addition to economic need. This study’s use of a nationally representative sample of African American, Black Caribbean and non-Latino White adults provides an important complement to the ethnographic and in-depth interview studies of middle-class White adults (Nelson, 2020), Haitian immigrants (Shaw 2008), older African Americans (e.g., Johnson, 1999), African American high school students (Alexakos et al., 2011), homeless African Americans (Desmond, 2012), and African Americans who reside in low-income communities (Stack, 1974).

The findings for the family, friend, and church member correlates indicate that fictive kin involvement does not result from a deficit in family contact and closeness (that is, low levels of family contact/closeness being associated with elevated fictive kin involvement). In fact, family, friend, and church member factors were all positively associated with our fictive kin outcome variables, suggesting that they may work together in a synergistic manner to facilitate fictive kin processes. Collectively, our findings are consistent with the notion of kinship extension as a means to strengthen family bonds. Additionally, our analysis demonstrates the presence of within-group variability in the demographic and social correlates of the prevalence, number, and supportive functions of fictive kin among all three populations under investigation. Taken together, this study is an important step forward in understanding the overall prevalence and demographic and social correlates of fictive relationships of African Americans, Black Caribbeans, and non-Latino Whites.