| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
a Purdue University, West Lafayette, IN
Correspondence: Janet M. Wilmoth, PhD, Department of Sociology and Anthropology, Purdue University, 1365 Stone Hall, West Lafayette, IN 47907-1365. E-mail: wilmothj{at}sri.soc.purdue.edu.
Vernon L. Greene, PhD
 |
Abstract |
|---|
 TOP Abstract Determinants of Immigrant Living...MethodsResultsDiscussionReferences |
|---|
Key Words: Residential dependence • Immigrant groups • Cultural preferences
Between 1990 and 2050, the older population in the United States will become increasingly diverse, with the proportion of the older population that is non-White projected to increase from 10.2% to 21.3% (Taeuber 1990
). The growth in the non-White segment of the older population is fueled not only by the aging of younger cohorts born in the United States but also by immigration. Over the last 40 years, immigration streams have shifted toward Latin American and Asian countries (Muller and Espenshade 1985). The basis for entry has also changed, with preference being given to family members of immigrants and those with desirable occupational skills or investments. These changes are not only increasing the ethnic diversity of younger immigrant cohorts that will grow old in the United States but have also increased the number of older immigrants (Treas 1997; U.S. Immigration and Naturalization Service 1993).
Despite the growing number of older immigrants in the United States, there are few studies that explicitly examine living arrangements within this group. Living arrangements among older immigrants are important to understand because each immigrant group has cultural preferences and individual-level characteristics that can potentially influence residential independence in later life. Previous research on the entire older population has shown that living arrangements influence the well-being of older adults and their family, while shaping reliance upon informal and formal social supports. For example, independent living arrangements—living either alone or with a spouse—are considered the most desirable for older adults in the United States because they offer more autonomy. However, these living arrangements (in particular living alone) can increase social isolation and reliance upon formal social supports. Because all individuals aged 60 or over in the United States are eligible for community services according to Title III of the Older Americans Act, understanding the characteristics of older adults who live independently and need assistance is essential to governmental policy formation and program planning.
Studies of ethnic living arrangements provide some insight into immigrants' living arrangements (Angel, Angel, Lee, and Markides 1999; Burr and Mutchler 1992, Burr and Mutchler 1993a, Burr and Mutchler 1993b; Himes, Hogan, and Eggebeen 1996; Kamo and Zhou 1994). These studies suggest that immigrants who arrived more recently, are less acculturated, and have fewer economic resources are more likely to live in multigenerational households. Furthermore, there are substantial racial and ethnic differences in the propensity of older adults to live in multigenerational households. Yet most of these studies provide limited information about immigrants because they are primarily concerned with explaining ethnic, not immigrant, living arrangements.
A few studies explicitly compare the living arrangements of immigrant and native-born older adults in the general population. In an analysis of older female immigrant living arrangements in Canada, Boyd 1991 found that immigrant women are more likely than native-born women to have extended family arrangements in later life. In addition, there are differences in the propensity to live with family among immigrants. A key defining factor is age at immigration: Women who immigrated at older ages are more likely to have extended family living arrangements. These later life immigrants tend to have characteristics that increase the likelihood of extended family living arrangements, such as lower levels of education and income, poorer English or French speaking skills, and more children.
Similar to Boyd 1991 findings in Canada, Wilmoth, De Jong, and Himes 1997 found that immigrants in the United States, particularly those aged 60 or older upon arrival, are more likely to live in extended family arrangements and less likely to live independently than older nonimmigrants. Furthermore, the differences between immigrant and nonimmigrant living arrangements are not due to differences in population composition, economic resources, functional limitations, or acculturation (Wilmoth et al. 1997).
Because previous research on older immigrants in the United States has been primarily concerned with either explaining the living arrangements of specific ethnic groups or comparing immigrants to nonimmigrants, it does not identify the individual-level characteristics that influence living arrangements among older immigrants. Further, it has not systematically documented living arrangement differences across and within particular immigrant groups. These differences are important to consider because each immigrant group has unique demographic characteristics, life experiences, and cultural beliefs that can potentially influence later life living arrangements.
Therefore, the purpose of this research is: (a) to identify the distribution of living arrangements across and within older non-Hispanic White, Hispanic, and Asian immigrant groups, taking into account differences by gender and marital status; and (b) to determine whether differences in living arrangements across these groups can be attributed to cultural preferences or other individual-level factors including acculturation, resources, needs, and availability of children.
|
Determinants of Immigrant Living Arrangements in Later Life |
|---|
|
TOPAbstractDeterminants of Immigrant Living...MethodsResultsDiscussionReferences |
|---|
Preferences are often inferred through measures of race and ethnicity. Numerous studies have documented variation in the living arrangements of older Americans by race and ethnicity. It is widely accepted that older Hispanic and Asian Americans are more likely than non-Hispanic Whites to live with family (Burr and Mutchler 1992, Burr and Mutchler 1993a; Himes et al. 1996; Kamo and Zhou 1994). The distinct living arrangements of minority elders are most often attributed to economic need and cultural traditions that influence living arrangement preferences (Angel and Tienda 1982; Burr and Mutchler 1993b; Himes et al. 1996). However, factors related to acculturation and immigration also influence minority living arrangements. Studies on minority elders' living arrangements that control for acculturation and immigration status consistently indicate that foreign-born individuals, particularly those who are less acculturated and have immigrated more recently, are significantly more likely to be living with family members in later life (Angel, Angel, McClellan, and Markides 1996, 1999; Burr 1992; Burr and Mutchler 1992, Burr and Mutchler 1993a, Burr and Mutchler 1993b; Kamo and Zhou 1994).
The influence of acculturation and immigrant status on later life living arrangements can be partially explained by immigrants' experiences. Immigrants brings living arrangement preferences, often preferences for extended family living arrangements, with them from their country of origin. While cross-cultural comparisons of elderly living arrangements are not common, there is evidence that multigenerational households are more common among older adults in Central and South America, as well as Asian and Eastern European countries, than in Western Europe or North America (De Vos 1990; Martin 1988; Wall 1989; Wolf 1990, Wolf 1995). The prevalence of multigenerational households in these countries can be attributed to cultural beliefs regarding reciprocity within the family, norms pertaining to the appropriate living arrangement of older adults, and the availability of housing.
For example, the structure of households in the northwestern part of Europe, as well as North America, has historically been focused on the nuclear family. Cultural norms in those countries that advocate self-determination, independence, and individualism encourage the increased residential privacy that is offered in nuclear households (Simic 1990). As a result, Scandinavian countries have the highest percentage of older adults living alone, followed by other northwestern European and North American countries. In contrast, extended multigenerational households are more common in Eastern, Central, and Southern Europe, where the norms regarding individualism are not as strong and family members are more economically interdependent. Countries in these regions of Europe tend to have a relatively low percentage of older adults living alone (Wall 1989; Wolf 1995).
Hispanic culture, which is prevalent in Central and South American countries, is characterized by strong kinship bonds, frequent intergenerational interaction, mutual exchanges between extended family members, and hierarchical family relationships (Becerra 1988; Sanchez-Ayendez 1988; Szapoczink and Hernandez 1988). These family-oriented cultural values create normative obligations to other family members, particularly older adults (Markides and Mindel 1987), which can increase likelihood of extended family living arrangements. In fact, previous research suggests that a majority of older adults in Latin America do live in extended family households (De Vos 1990).
The situation is similar in Asian countries, where approximately 75% of older adults live with their children (Martin 1988). However, the cultural values driving extended family living arrangements in this region of the world are different. Buddhism and Confucianism, in conjunction with rural folk religions, tend to increase the status of older adults in many Asian countries. The practices of ancestor worship, filial piety, and primogeniture encourage Asian youth to revere and respect older adults (Kiefer 1990; Olson 1990). Individuals are encouraged to subvert personal needs and desires in the interests of larger groups, including the family, and there are strong norms of reciprocity (Akiyama, Antonucci, and Campbell 1990; Kiefer 1990). Multigenerational households are common because these cultural traditions encourage elderly parents to live with children, primarily the oldest son. In urban areas, where exposure to Western values should place downward pressure on such extended family living arrangements, shortages in housing supplies and limited state-supported social services maintain a high level of multigenerational households (Logan, Bian, and Bian 1998; Olson 1990).
As this discussion indicates, immigrants from various regions of the world should arrive in the United States with culturally based preferences for extended family living arrangements. The strength of this preference should vary across different immigrant groups. Once in the United States, national immigration policies that encourage family reunification reinforce country-of-origin living arrangement preferences by requiring sponsors to sign nonbinding affidavits of support. Together, individual preferences rooted in specific cultural traditions and national immigration policies foster family living arrangements among immigrants by increasing family-oriented obligations (Boyd 1991; Wilmoth et al. 1997). However, acculturation should decrease the likelihood of living with family.
The increased likelihood of living with family among immigrants may not be due to preferences, but other individual-level characteristics, such as resources, needs, and availability of children. Previous research consistently shows that economic resources facilitate independent living arrangements because these resources enable older individuals to purchase preferred living environments as well as services that maintain independent living (Burr and Mutchler 1993a; Soldo, Wolf, and Agree 1990). For minorities and immigrants, extended family living arrangements can be seen as a strategy to pool income and other resources, particularly at key points in the life course (Angel and Tienda 1982; Blank 1998; Blank and Torrechilha 1998; Worobey and Angel 1990a). Thus, extended family living arrangements among older adults are not necessarily initiated because of the needs of the aging individual. Family finances or other circumstances, such as the required care of a child or older adult, can prompt the formation of multigenerational households.
In addition to economic resources, health status also influences living arrangements in later life. Declining health decreases the likelihood of independent living arrangements while increasing the likelihood of dependent living arrangements (Angel 1991; Avery, Speare, and Lawton 1989; Mutchler and Burr 1991). Overall, however, economic resources are believed to have the most influence on later life living arrangements. Independent living arrangements are often maintained among older adults with more economic resources even when their health fails (Soldo et al. 1990; Worobey and Angel 1990b).
Number of children is often used as a measure of the potential pool of people with whom an older adult can coreside. Overall, number of children is negatively related to the likelihood of living alone; older adults without children or with only one child are significantly more likely to be living alone (Avery et al. 1989; Soldo et al. 1990). Women with daughters only are more likely to be living with a child than women with only sons, while women with both sons and daughters have intermediate probabilities of living with a child (Soldo et al. 1990). Although these studies do not specifically consider how the availability of children affects immigrant living arrangements, it is reasonable to expect similar results among immigrants.
Finally, it is important to note that later life arrangements are systematically related to gender and marital status. Married older adults are more likely to be living only with a spouse, while unmarried older adults are more likely to be living alone. Furthermore, women are more likely to live alone or with family than men because of gender differences in mortality (Siegel 1993). As a result, many studies of older adult living arrangements (e.g., Boyd 1991; Burr and Mutchler 1992; Choi 1991; Soldo, Sharma, and Campbell 1984; Wolf and Soldo 1988) specify separate models based on gender and marital status.
To summarize, it is expected that the distribution of living arrangements will vary across different immigrant groups. In particular, Hispanic and Asian immigrants are expected to be more likely to live with family than White immigrants—especially White immigrants from North America and Western Europe. Furthermore, variation is expected within each immigrant group due to specific country-of-origin cultural beliefs and practices. It is expected that some of this variation across and within groups will be accounted for by individual-level characteristics including acculturation, resources and needs, and availability of children.
|
Methods |
|---|
|
TOPAbstractDeterminants of Immigrant Living...MethodsResultsDiscussionReferences |
|---|
The citizenship and year of entry into the United States items are used to determine immigrant status. The only respondents included in the analysis are individuals who were citizens of another country at birth and subsequently moved to the United States (n = 63,953). All respondents born in the United States or overseas to citizens of the United States are considered nonimmigrants and thus are excluded from the analysis.
Similar to previous research (e.g., Burr 1992; Burr and Mutchler 1992; Mutchler and Burr 1991), living arrangements are operationalized in terms of complexity and householder status. For this analysis, immigrants could be in one of four living arrangements: alone, with spouse only, with family as the householder (i.e., the person listed in column 1 of the census form who is generally considered to be the head of the household), and with family not as the householder. Living alone and living with a spouse only are collapsed into a single reference category called independent living arrangements. Identifying whether an older adult living with family is the householder is an important indicator of dependence (Burr 1992; Burr and Mutchler 1992; Mutchler and Burr 1991). According to the U.S. Census Bureau definition, the householder is either the homeowner or the renter of the property, which suggests the individual is economically independent. For this analysis, women who are married to the householder are also classified as a householder. Being the householder, or married to the householder, implies access to and control of more economic resources. Therefore, living with family as a householder suggests an interdependent living arrangement, and living with family without being the householder can be considered a more dependent living arrangement.
Although living in an institution is the most dependent living arrangement, institutionalized respondents are excluded from the analysis for two reasons. First, rates of institutionalization are low for immigrants (Wilmoth et al. 1997). Second, because information on personal income is not available for institutionalized respondents and income is related to the chance of institutionalization, it is not possible to correctly specify a model for this group. Immigrants living with nonrelatives are excluded from the analysis because there were not a sufficient number to identify immigrant group differences by gender and marital status.
One of the strengths of the PUMS data is that there are enough cases in the most common living arrangement categories to provide a detailed comparison of immigrant groups. The questions on self-identified race, Hispanic origin, place of birth, and ancestry are used to identify 11 immigrant groups: three non-Hispanic White groups (from North America or Western Europe, Eastern Europe, and other countries), four Hispanic groups (from Mexico, Central and South America, Cuba, and other countries), and four Asian groups (from China or Taiwan, Japan, India, or other countries). Blacks (from African or other countries) and Native American Indians (from Canada or Central and South America) are excluded from the analysis because neither group has enough immigrants to support a multivariate analysis. These immigrant group measures capture the effect of country-of-origin cultural preferences for particular living arrangements within three general ethnoracial groups.
The analysis includes three items that control for acculturation. The first is a measure of age at immigration. This item is constructed from age in 1990 and the year of immigration item. Because the year of immigration variable is categorical, the upper and lower bounds of the year categories were used to determine the potential range of ages within which the respondent immigrated. The mean of the age range was used as the age at immigration. For this analysis, the age at immigration was collapsed into a dichotomous variable that controls for immigration at age 60 or older.
The second measure of acculturation controls for the year of immigration. This variable includes four categories: prior to 1965 (reference category), 1966–1974, 1975–1984, and 1985–1990. Immigration prior to 1965 is used as the reference category because 1965 marks the point at which migration streams shifted toward Latin American and Asian countries. This variable not only controls for this shift in migration streams but also how recently the respondent immigrated. It is expected that the more recent the immigration, the less time the respondent has had to be acculturated. Subsequently, more recent immigrants should be more likely to live with family. Although this variable will serve as an important control, it will not be possible to distinguish between the effects of entry time period and the effects of years since entry, because to do so would require at least two cross-sections of data (Borjas 1985).
English language proficiency is the final measure of acculturation. Respondents who are unable to speak English or who speak English poorly are coded as "poor," and respondents who speak English well, very well, or only speak English are coded as "good" (reference category). Being able to speak English proficiently is not only necessary for independent living but it is also a primary means of exposing immigrants to the dominant culture. Individuals who have better English language skills should be more acculturated and thus less likely to live with family in later life.
Personal income, education, and functional limitations serve as the measures of resource and need factors. Personal income is measured continuously in U.S. dollars. Education is coded to have three categories: high school or less, high school graduate (reference), and some college or more. The functional limitation variable was taken from the questions: "Because of a health condition that has lasted 6 or more months, does this person have any difficult (a) going outside the home alone, for example, to shop or visit a doctor's office? Or (b) taking care of his or her own personal needs, such as bathing, dressing, or getting around inside the home?" The recoded variable in this analysis has three categories: having no limitations (reference category), having a mobility or personal care limitation, having both mobility and personal care limitations. While these measures of functional health are rather crude, they are the only measures of functional limitations available in the PUMS data.
Unfortunately, PUMS has limited information regarding the availability of children. Only female respondents provide information regarding the number of children ever born, and there are no questions concerning the gender composition of children or proximity of children. Thus, this analysis can only include a limited control for availability (i.e., number of children) in the models for unmarried and married women. This is an important limitation of these data and must be considered when interpreting the results.
These measures are first used in a descriptive analysis that focuses on the distribution of living arrangements among the immigrant groups by gender and marital status. Then, multinomial logistic regression models are presented that predict living arrangements. Separate models are specified by gender and marital status. The reference category of each model is independent living arrangements (i.e., living alone or with a spouse only). The descriptive statistics and multivariate models will be presented together for each gender/marital status group: unmarried men, married men, unmarried women, and married women. Then, similarities and differences across these four groups will be summarized.
The purpose of the multivariate models is to determine whether controlling for individual-level characteristics accounts for the observed differences across the immigrant groups. If controlling for these variables eliminates the significant differences across the immigrant groups, it will imply that observed differences in living arrangements are due to variation in individual-level characteristics. Conversely, if these controls do not influence the differences across immigrants groups, it suggests that cultural preferences influence immigrant living arrangements.
Partial models that sequentially entered the acculturation, resource, need, and availability measures were estimated to identify which set of variables account for group differences. The magnitude of the coefficients did not vary enough across these models to justify presenting the partial models in the tables (partial models are available from the author upon request). The few differences between the partial and the full models are noted in the text.
|
Results |
|---|
|
TOPAbstractDeterminants of Immigrant Living...MethodsResultsDiscussionReferences |
|---|
|
|
|
Married Men.
The second panel of Table 2 , which presents the results for married male immigrants, reveals that living independently is most common among those who are White, Japanese, or Cuban. For the rest of the immigrant groups the most common living arrangement for married men is with family as the householder. It is interesting to note that for married men the highest rate of living with family without being the householder (9.9%) is among Asian Indians.
The second column in Table 3 presents the multivariate models for married male immigrants. All of the groups, except Whites from Eastern Europe and Asians from Japan, are more likely to live with family even after controlling for other characteristics. Married Japanese men are only more likely to live with family as the householder. Overall, for married men, the likelihood of living with family as the householder is highest among Asians from India and other countries, as well as Hispanics from Mexico and Central/South America. These groups also have the highest risk of living with family without being the householder.
Unmarried Women.
The distribution of living arrangements among unmarried female immigrants is similar to that of unmarried male immigrants; however, unmarried women are generally less likely to live independently and more likely to live with family without being the householder. White, as well as Japanese, immigrants are the most likely to live independently. Living with family without being the householder is more common among Hispanic and Asian immigrants. However, there is considerable variation within each of these groups—particularly among Asian immigrants. For example, 85% of unmarried female immigrants from India live with family without being the householder, compared to only 28.6% of those from Japan. Interestingly, the highest rates of unmarried women living with family as the householders are among Hispanic groups, with the percentage ranging from 17.7% among Cuban immigrants to 25.3% among Mexican immigrants.
The first column of Table 4 presents the multivariate model for unmarried female immigrants. All of the immigrant groups are more likely to be living with family, regardless of householder status, except two—Whites from Eastern Europe and Asians from Japan. Partial models (not shown) indicate that the coefficients for Japanese immigrants become nonsignificant after controlling for acculturation and number of children. Therefore, living arrangement differences between unmarried female Japanese immigrants and the reference group can be attributed to differences in acculturation and availability of children. Overall, among unmarried female immigrants, the risk of living with family is highest among Asians from India and other countries, as well as Hispanics from Central or South America.
|
The second panel of Table 4 presents the model for married female immigrants. Whites from Eastern Europe and Japanese immigrants are not significantly different from the reference group. Whites from other countries are only more likely to be living with family without being the householder. All other groups are significantly more likely to be living with family regardless of householder status. Overall, the highest risks of living with family are among Asian immigrants from India and other countries, followed by Hispanic immigrants from Mexico and Central/South America.
Summary
Overall, Table 2 demonstrates the considerable variation in living arrangements across the immigrant groups. Among unmarried male and female immigrants, the distribution of living arrangements among Whites from North America or Western Europe is significantly different from the distribution among all the Hispanic and Asian immigrant groups. Among married male and female immigrants, the living arrangement distribution among Whites from North America or Western Europe is significantly different from Whites from other countries as well as all the Hispanic and Asian immigrant groups.
Keeping this variation in mind, rates of dependent living arrangements (i.e., living with family without being the householder) are highest among unmarried female immigrants. Rates of interdependent living arrangements (i.e., living with family as the householder or married to the householder) are highest among married males and females. Regardless of gender or marital status, the highest rates of independent living arrangements (i.e., alone or with a spouse only) are among White, Japanese, and Cuban immigrants.
These observed differences in living arrangements could be due to variation in individual-level characteristics. Therefore, multinomial logistic regression models were tested that include controls for acculturation, resources and needs, and availability of children. Table 3 and Table 4 indicate that controlling for these variables does not explain the increased risk of living with family among the various immigrant groups.
In terms of the other variables in the model, acculturation generally decreases the likelihood of family living arrangements. Migration at age 60 or older decreases the likelihood of living with family as the householder for married men and unmarried women, while increasing the likelihood of living with family without being the householder for married men, unmarried women, and married women. Recent immigration and poor English language proficiency increase the risk of living with family.
Limited resources and high needs tend to increase the likelihood of family living arrangements. Specifically, income tends to decrease the risk of living with family for men and women, with two important exceptions. Income increases the risk of living with family as the householder for unmarried women but was not a significant predictor of living with family as the householder for unmarried men. Immigrants with less than a high school education are not more likely to be living with family than immigrants who are high school graduates. However, having a college education decreases the likelihood of living with family as the householder for married men and decreases the likelihood of living with family without being the householder for unmarried as well as married men. Finally, physical limitations decrease the risk of living with family as the householder for married men and women, but increase the risk of living with family without being the householder for unmarried men and women.
The effect of child availability could only be assessed in the models for women. Regardless of marital status, number of children increases the likelihood of living with family for female immigrants. Most likely, number of children is also influential for men, but the nature of this relationship can not be determined because of data limitations.
|
Discussion |
|---|
|
TOPAbstractDeterminants of Immigrant Living...MethodsResultsDiscussionReferences |
|---|
points out, "Although there is a substantial literature on the problems of immigrants, the issues faced by older immigrants have been for the most part neglected" (p. 367). This research has attempted to address this gap in the literature by examining differences in living arrangements among older immigrants. Overall, the findings suggest that the factors influencing living arrangements among the general population in the United States also influence living arrangements among the immigrant population. Similar to the findings of previous research (e.g., Burr and Mutchler 1993b; Himes et al. 1996; Wilmoth et al. 1997), later life living arrangements among immigrants are influenced by racial and ethnic characteristics, which are generally considered indirect measures of cultural preference for particular living arrangements in later life. However, immigrants have the added influence of acculturation. Immigrants who have arrived relatively recently, were age 60 or older upon arrival, and who have poor English language skills are the least likely to have residential autonomy in later life.
Although some of the acculturation, resource, need, and availability measures are significant, controlling for these characteristics did not explain the group differences in living arrangements. The one exception was for unmarried female Asian immigrants from Japan. Controlling for acculturation and, in particular, number of children did explain the higher rates of living with family among this group. However, the inability of these controls to eliminate the majority of the group differences suggests that preferences, which are shaped by the immigrant experience as well as cultural background, influence immigrant living arrangements in later life. Although it is not possible to directly measure preference with census data, this finding is consistent with previous research (Himes et al. 1996; Wilmoth et al. 1997), which concludes that indirect measures of cultural preference are important determinants of later life living arrangements, particularly among minority immigrants. Preferences, along with resource, need, and availability factors, set the stage for later life living arrangements among this group.
This analysis also indicates that there are different propensities to live in multigenerational households across and within the various groups. Consistent with the findings of Himes and colleagues 1996 regarding minority elders, older Asian and Hispanic immigrants are more likely than non-Hispanic White immigrants to live with family. However, the findings also indicate that there is considerable variation within specific older minority immigrant groups. This suggests that combining respondents from different countries of origin might not be appropriate when studying the living arrangements of immigrants, particularly Hispanic and Asian immigrants.
While the risk of living with family was relatively similar among the non-Hispanic White groups, there was considerable variation within the Hispanic and Asian groups. Among the Hispanic groups, those from Cuba had the lowest risk of living with family. While this group exhibits strong kinship bonds, Cuban families in the United States started becoming nuclear more than 30 years ago (Szapoczink and Hernandez 1988). Among immigrants from Asia, Japanese immigrants have notably different living arrangements. However, there is evidence that Japan is different from other Asian countries on several accounts due to the relatively high standard of living and modernization. Compared to other Asian countries, multigenerational households are less common, and the percentage of older adults living with children has been declining (Martin 1988). Therefore, compared to other Asian immigrants, Japanese immigrants should arrive in the United States with weaker preferences for extended family living arrangements.
The PUMS data contain only one limited measure of availability of children. The number of children ever born is asked only of women, and there are no questions concerning the gender composition or proximity of children. This is a potentially important issue for model specification because previous research on the older population in the United States has shown that children decrease the likelihood of autonomous living arrangements in later life (Avery et al. 1989; Soldo et al. 1990). The models in this analysis indicate that the number of children significantly increases the chance of living with family for women, regardless of headship, but the magnitude of this effect is not great. More important, this variable does not significantly change the other coefficients in the model, including the immigrant variables (with the exception of the coefficients for Japanese women who are unmarried). Therefore, while availability of children is most likely an important predictor of living arrangements among immigrants, it does not appear to explain immigrant group variations. However, further research that is able to measure children's characteristics in more detail is required to confirm these findings.
In conclusion, these findings have important implications for the future distribution of living arrangements among the older population. It can not be assumed that the proportion of older adults living alone will continue to increase in the future. Given the documented differences in later life living arrangements among minority and immigrant groups, the distribution of living arrangements among future elderly cohorts could be considerably different from those of current cohorts. The proportion of older adults living in extended family arrangements could increase as minority and immigrant populations grow, particularly if immigrants continue to have limited access to social welfare programs and sponsoring families have to assume additional responsibility. Future research should monitor changes in the distribution of living arrangements among older immigrants and assess the impact these living arrangements have on the well-being of older immigrants and their families.
|
Acknowledgments |
|---|
Received for publication April 7, 1999. Accepted for publication May 23, 2000.
|
References |
|---|
|
TOPAbstractDeterminants of Immigrant Living...MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Diwan Limited English Proficiency, Social Network Characteristics, and Depressive Symptoms Among Older Immigrants J. Gerontol. B. Psychol. Sci. Soc. Sci., May 1, 2008; 63(3): S184 - S191. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. E. Firbank and J. Johnson-Lafleur Older Persons Relocating With a Family Caregiver: Processes, Stages, and Motives Journal of Applied Gerontology, April 1, 2007; 26(2): 182 - 207. [Abstract] [PDF]
|
|
|
|
|
|
S. L. Brown, G. R. Lee, and J. R. Bulanda Cohabitation among older adults: a national portrait. J. Gerontol. B. Psychol. Sci. Soc. Sci., March 1, 2006; 61(2): S71 - S79. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Wilmoth and P.-C. Chen Immigrant Status, Living Arrangements, and Depressive Symptoms Among Middle-Aged and Older Adults J. Gerontol. B. Psychol. Sci. Soc. Sci., September 1, 2003; 58(5): S305 - 313. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Kim and D. S. Lauderdale The Role of Community Context in Immigrant Elderly Living Arrangements: Korean American Elderly Research on Aging, November 1, 2002; 24(6): 630 - 653. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
