Older Adults’ Friends and Ethnicity

This study examines the relationships among friends and ethnicity of older adults. Friends includes friend numbers and their quality of relationships with friends of older adults in the current study. Data from the National Social Life, Health, and Aging Project (NSHAP) survey (Waite et al., 2020) were used. The NSHAP study sampled persons 57-85 years of age ( n =3005). The respondents completed a telephone survey in which they reported their background information (e.g., income, gender, race, age, health, retirement status, and marital status) and social network characteristics. It was hypothesized that older adults’ ethnicity differentially influenced family relations. In comparison to Anglo older adults, African and Hispanic older adults have weaker (smaller number and less cohesive) family culture. In order to identify the associations between ethnicity and friend relations, multiple regression analysis was used. Results revealed that African American and Hispanic older adults reported larger numbers of close friends, higher quality of friend in general, and higher frequency of contact with them compared to Anglo older adults. The current study’s findings build on a convoy model to account for how older adults’ ethnicity is differentially associated with their quality and size in friend relationship for future research is to examine more diverse in friend and ethnicity variables which explain the dynamic relationships between older adults’ demographic factors and friend network.


Introduction
The elderly population is one of the fastest growing populations. In 2050, it is estimated that the aging population occupied 12% of the population; it is expected to increase further to 20% of the population by 2050. The elderly population is also becoming more ethnically diverse. According to a U.S. Census Bureau Statistical Brief (2020), one in ten of the elderly were a race other than White in 1994; however, that proportion is expected to rise to two in ten by 2050. Naturally, there is a growing concern over how society will create the resources to deal with the needs of today's growing elderly population. Sources of social support among the elderly could be better understood, especially in the current environment of social welfare spending cuts that threaten the federally and state supported programs upon which many elderly depend.

Friend Network in Later Life
In reference to the older adults' friend networks, past research noted that having at least one friend may be critical to well-being in later life (Hooyman & Kiyak 2020). Roberto and Scott (1986) pointed out diverse types of friends (e.g., best friend, life-long friend, close friend) in which it is possible to maintain close ties despite infrequent contacts if older adults have shared significant past histories with the friends. Through the associations with friends, older adults can exchange comparable social resources and function as confidents to each other in later life.
Researchers (Feldman, 2013) cited three aspects of friend relationships in later life. First, individuals who are in the friendships may share similar age, socio-economic status, and ideas or life styles. Second, friends' relationships are based on mutuality, unlike family relationships (Grafanaki et al, 2005). Third, older adults can extend their social activities and daily lives with friends by exchanging help (Mason 2013). Allen, Lee and Bakk (2001) distinguished three main differences in friendships from family relationship such as lack of formal roles, voluntary, and non-exploitive nature. For example, Lennartsson and Silverstein (2001) emphasized that family may provide emotional and financial help whereas friends may offer mutual gratification, socialization, and a sense of reciprocity. Given the three functions of friend relationships of older adults, it is possible to have an opportunity for intimate discussion among the friend networks whereas older adults' may be more reluctant to discuss their intimate matters with their children. In addition, the voluntary characteristic of friend relationships may provide mutual help without feelings of obligation for emotional and material support when the older adults experience loss of spouse, retirement, and physical decline. For example, older adults who lost their spouse can understand their friends' situations better than adult children because the friend may have faced similar life events (Mason, 2013). Given the friend relationships characteristics in later life such as their voluntary nature, mutual support, and similar life stage and experience, the friend relationships may help individuals to age successfully by enhancing the sense of coherence and sympathy about older adults' lives (Olson, Defrain, & Skogrand, 2007). Jang, Mortimer, Haley, and Graves (2004) noted that social support involves aid, affection or affirmation in formal or informal contexts. As an informal social support, Hillier and Barrow (2015) found that children and friends are important informal support networks for both poor and higher income older adults. With 571 rural older adults, canonical correlation analysis revealed that proximity to children and friends was a factor for networks of both of poor and higher income groups. In addition, Montenegro (2020) noted that family and friends' exchanges are different across the life course. Similarly, The Centers for Disease Control and Prevention Survey (2022) suggested that family support might have a source of negative exchange (e.g., criticize, demand) when individual experience marital changes (e.g., divorce, widowhood) whereas friend relationships have a more positive impact.
Although previous research indicated that friendship may offer information and companionship (Jaumot-Pascual, Monteagudo, Jleiber, & Cuenca (2016), Mason (2013) found that elderly adults' feelings of bereavement on the loss of a close friend are similar to the experience of loss of a family member. Therefore, it is clear that older adults' friendship may provide psychosocial well-being, informal support, and successful adaptation to aging (Hooyman & Kiyak, 2020) In reference to the relationships between demographic factors and friend relationships in later life, Feldman (2013) noted that quality of friend networks may associated with older adults' financial resources, health, and marital status which accompany with older adults' daily lives. Fisher et al. found that there might be envy or jealously within friend relationships which might cause less frequent association with their friends or lower quality of their friend networks. Specifically, if there is difference in their financial ability, marital status, and health status, the lower position of older adults may feel envy toward the higher position of their friends. Similarly, in the study of Ungar and Florian (2004) revealed that availability of friendships in later life is related to older adults' marital status, health, and income level. Specifically, due to the friend relationships' couple-oriented nature, availability and frequency of contact with friends in later life might be influenced by older adults' marital status. In addition, older adults' poor health and lower financial ability might cause limited contact or interaction outside home and other social activities with friends. Therefore, it is clear that older adults' marital status, health, and income factors might have an influence on the friend networks. However, it is still needed to define to what extent older adults' marital status and retirement predicts the size and quality of the friend networks. In addition, further research is required that examines how much older adults' friend networks can be influenced by income, health, gender, race, and age factors.

Ethnic Diversity
The aging population may share commonalities in regards to the types and functions of their friend networks, but in other ways, the aging population is very diverse. Different ethnic groups are disproportionally represented among the aging population. For instance, White older adults represent 84% of the total elderly population in the United States (67.1 million of 55 and older adults), African American older adults represent 8% of the total elderly population, Hispanic older adults represent 6% of the total population, and "other" ethnic groups (e.g., Asian, Pacific Islanders, American Indians) represent 3% of total population with a decline in White (U. S. Census Bureau, 2020). In 2050, it is projected that older African Americans will increase to 12.2% of the total elderly population, and older Hispanics will represent 16.4% of that population. In reference to the Asian/Pacific Islanders and the American Indian older adult population, Chinese represent 30%, Japanese represent 24%, Filipino represent 24%, Korean represent 8%, American Indians represent 5%, and other ethnic groups represent 5% of that particular older adult population (U. S. Census Bureau, 2020). The minority population is of adult citizens on the rise.
The Asian/Pacific Islander population is the fastest growing among aging ethnic minority groups due to their increasing rates of immigration (U.S. Census Bureau, 2020). This rapid growth in the aging Asian/Pacific Islander minority population calls for greater focus on understanding their particular needs. Cultural studies can provide insight into differences among ethnic aging populations. For instance, older Asian/Pacific Islanders and American Indians share the value of familism, which defines family as central to the functioning of the family. (Hooyman & Kiyak, 2020). The older populations of these ethnic groups hold strong family-centered beliefs. For example, they believe that older parents should be cared for by adult children and should also be respected as the leader of the family or household. However, many of these family-oriented values have been eroded as a result of reductions in family size, increased mobility among family members, work-related family life, and acculturation into the larger society (Hooynam & Kiyak, 2015).
Ethnicity may also be a variable that can help to explain some differences in friend networks of older adults. Socioeconomic status, education and health often vary among different ethnic groups. A review of the literature on res.ccsenet.org Review of European Studies Vol. 15, No. 2;2023 ethnic differences among older adults provides insight. For instance, in comparison to white older adults, African American and Hispanic elders have lower economic status; lower levels of education and poorer physical health (Mason, 2013). Lower socioeconomic status appears to be a significant variable in friend relations. Olson and Skogrand (2007) noted that friend relationships are related to an individual's social status in which middle-class older adults are more likely to have larger numbers of friends than those of older adults with lower socioeconomic status. Similar findings have revealed that older African American and Hispanic adults are less likely to engage with their friends than White older adults because the former spend more time with their large extended families, which consumes a good deal of their time (Feldman, 2013). These findings further indicate the need to understand cultural differences.
Ethnic differences are also found in the number of friends older adults have. Hiller and Barrow (2015) found that being Black was associated with smaller social networks. Additionally, a study of older adults residing in San Francisco, California found that White respondents consistently reported a larger number of friends than the African American and Hispanic respondents, which was linked to the Caucasian group's higher socio-economic status (Lee & Bakk, 2001). However, the way in which an ethnic group classifies or counts friends in their network can influence the size of an elderly group's friend network. According to Hooyman and Kiyak (2020), African Americans may have a more flexible definition or concept of friend networks than Whites. African Americans often see close friends and neighbors as "fictive kin", forms of social ties that are based on neither blood ties nor by marriage ties (National Council on Aging, 2021) as they exchange support and provide help (Hooyman & Kiyak, 2020). Similarly, Montenegro (2020) found that African Americans considered relatives, neighbors, and close friends as part of their family networks, in which they have a large potential pool of social relationships.
The life course perspective guides the researcher to think about how diverse the aging experience can be based on differences in gender, ethnicity and social class. Ungar and Florian (2004) explained different worlds of aging to describing the different realities older adults experience based on their gender, ethnicity and social class and the inherent disadvantages they have experienced because of these variables.
The present study sought to fill in some of the gaps in the literature on older adults from different ethnic groups and their friend relationships. Given the inevitable life transitions in later life and the growing population of ethnic minorities, it is important to examine the associations of ethnic factors on older adults' friend networks. The purpose of the study was to gain more knowledge about the differences among older adults from different ethnic groups in relation to their friend networks. Based on the review of the literature of friend networks of older adults, the following hypothesis guided the study: Caucasian older adults have a stronger friend network than ethnic minority older adults.

Sample
An experimental design was employed using secondary data from a national data set. The National Social Life, Health, and Aging Project (NSHAP, see Waite et al., 2020) was used for the secondary data analysis in this study. The NSHAP is a national project funded by the United States Department of Health and Human Services, which examines older adults' health and social factors with a survey. Within the data set, information on the quality and quantity of older adult's friend networks was gathered for this study.
The researcher obtained permission from the University of Michigan to utilize the secondary data set. The data collected is representative of older adults as it is a nationwide, randomly selected sample. The sample comprised older adults whose ages ranged from 57-85 years. The mean age of the sample was 69.3 (SD=7.9). Forty eight percent of the total sample (n=3005) was male and fifty two percent was female. Ethnicity was assessed via single-item questions. Self-identified ethnicity was measured as a nominal level of measurement (e.g., 1=" White/Caucasian", 2=" Black/African American", 3= "Hispanic or non Black", 4= "Other"). The ethnicity composition was Caucasian 70%, African American 17%, Hispanic 10%, and other ethnicity 2.3%. (see table 1)  (Note. Frequency of friend demands total R 2 =.011, Frequency of friend criticize total R 2 =.013, p<.001, *p<.05 ** p<.01 ***p<.001, reference groups were married, retired, White, female, and middle-old group).

Discussion
The current study examined the friend networks of a group of older adults from different ethnic backgrounds. The findings indicated that there was a relationship between ethnicity and older adults' friend networks. Among the ethnic groups studied, African American older adults reported smaller sizes of friend relations and also experienced demands and criticism from their friend networks more frequently than the Caucasian group. Similarly, Hispanic participants in the study reported a smaller size friend relations and that they relied on friends less frequently than the Caucasian participants. These finding are consistent with the findings in previous studies (Jang, Haley, and Graves, 2004). For instance, Lennartsson and Silverstein (2001) explained how ethnic differences play a role in the older adult's help-seeking social behaviors. Among the samples of adults who provided care giving for an older family member diagnosed with dementia, the Hispanic ethnic group reported a smaller size friend network and a smaller size pool of available support outside the family (e.g., friend); the network was smaller than the Euro-American group given the same circumstances.
Another study of caregiving behaviors among different ethnic groups noted that a group of Hispanic participants were less likely to use non-family social support, such as friends, when needed, than a Caucasian group of participants. The researchers postulated reasons for their findings, one being that the Hispanic group prefers to keep the exchange of support within the family, due to the unique cultural mechanisms such as filial obligation (Bennet, Smith, & Hughes, 2005). The results of the current study led the researchers to concur with the idea that the cultural values often held by Hispanic groups, such as Familism and valuing family tradition, may help to explain why Hispanic participants rely on friends less frequently. In comparison to the Caucasian participants in the current study, the African American and Hispanic participants were more likely to rely on family, where the Caucasian participants were more likely to use a diverse source of social networks.
The current study also revealed some interesting findings in regards to ethnic differences, and negative experiences within the friend networks of older adults. The findings are again consistent with the literature. Grafanaki, Pearson, Cini, Mckenzie, and Anderegg (2005) noted that minority groups such as African Americans, Hispanics, and other ethnic groups have lower levels of socio-economic status (e.g., income, health, education) than Caucasians, which may be linked to a limited use of friend networks. In addition, African American and Hispanic groups of older adults' poor health status may cause friends in their network to experience care giving support burden (Jaumot-Pascual, Monteaudo, Jleiber & Cuena, 2016). One can assume that there may be a higher chance of demands and also criticism on the friend relations when the older adult has fewer resources (e.g., income, available social members), and his or her health declines. It is also possible to assume that older adults of higher socio-economic status may maintain positive friend networks and may have fewer negative experiences, such as demands or criticisms from their friends. Reflecting on the nature of friend relationships later in life (e.g. voluntary and provide for the exchanging of mutual support), one might suppose that African American and Hispanic groups of older adults may feel difficulty maintaining the relationships if one side of the friend relationship has a lower position in health and income status.

Future Directions
The current study provided a sketch of older adults' friend networks and the similarities and differences among different racial groups. Understanding the effects of different ethnic factors on the friend networks of older adults may aid gerontological researchers, family policy makers, and social welfare workers in designing social programs which provide the most benefits to older adults and their family, friends, and other social members.
Although not measured in the current study, a review of the previous research on ethnicity and friend networks in later life, indicates that African Americans and Hispanics have similar or larger sizes of friend networks, but their quality of friend networks might be lower than the Caucasian group's due to their lower socio-economic status and lower usage of social services (Pagano, Barkhoff, Heiby, & Schlicht, 2006) . In addition, because African Americans and Hispanics tend to highly value family life and tradition, their friend networks might be less strong (e.g., smaller and lower quality) than those of Caucasians. Additional variables such as health, income, education and work experience should be included in future aging studies. Longitudinal studies could also prove beneficial. The findings from the current study and previous research highlight the importance of Walsh, Green, Holahan, Cance, and Lee's (2019) idea of "different worlds of aging" and the need to do more research to understand the different realities of older adults based on their cultural experiences. Further exploration of cultural theories as related to aging may help expand the understanding and ability to meet the needs of a growing, diverse, aging population.
There might be limitations exist that influence the interpretation of the results. First, this research was a secondary data analysis of the NSHAP (2020) interview study; so, it is inherently restricted to the design of the original study. Therefore, it was not possible to refine the original interview questionnaire protocol. In addition, the variance of the results is meaningful in that current study used secondary data set which has the huge total sample (n=3005) size. Second, all of the variables were measured with one-item questions. Third, in reference to the social networks, a limited range of friend network variables were assessed, such as both size and quality of friend networks. The limited measure of friend networks entails that the results might have been different if additional social networks (e.g., volunteer groups, other community groups) were measured. Therefore, there needs to be a discussion of the appropriateness of regression with limited variables. Fourth, the findings are based on self-reported measures of demographic variables and friend networks, which may result in inaccuracies in reporting. Fifth, a cross-sectional study of this type can only reveal associations among variables; therefore, causality cannot be proved.
Results from the current study have implications for future researchers. They need to expand the number of variables measured. There might possibly be other factors that may mediate the older adults' friend networks which need to be examined for future studies. Therefore, future studies will find the direct and indirect associations between demographic factors and other social networks that are not investigated in the current study. Further studies also need to investigate the possible barriers that are related to demographic factors' influences in older adults' friend networks. By applying longitudinal design, it is possible to determine how the effects of friend networks and ethnic differences change over the life cycle.