Neighborhood Characteristics, Neighborhood Satisfaction, and Loneliness Differences Across Ethnic–Migrant Groups in Australia

Abstract

Objectives

Loneliness is an important concern for older adults. Studies have linked demographic characteristics with loneliness, showing that it varies by ethnic and migrant statuses in countries in Europe and North America. Characteristics of the physical environment in which older adults are embedded have also received some attention, though prior studies have not fully investigated whether older adults from different ethnic–migrant backgrounds may report variation in loneliness because of characteristics of, or satisfaction with, their neighborhoods, which may shape their social interactions.

Methods

Drawing on up to 4 waves of data from the Household, Income, and Labour Dynamics in Australia Surveys and random-effects models, I examine whether loneliness differs across ethnic/migrant groups in the Australian context. Furthermore, I examine whether neighborhood characteristics (its conditions and sociality) and satisfaction with the neighborhood (with its safety, with the neighborhood itself, and with “feeling part of the local community”) may be mediators for the association between ethnic–migrant status and loneliness.

Results

Findings show migrants from non-English-speaking countries report higher levels of loneliness, as compared with native-born, non-Indigenous Australians. More favorable neighborhood characteristics and higher levels of satisfaction with different aspects of the neighborhood are consistently associated with lower levels of loneliness. Neighborhood sociality and satisfaction with aspects of the neighborhood partially mediated the association between ethnicity status and loneliness for migrants from non-English-speaking countries.

Discussion

This study showed loneliness differs across older Australians of different ethnic and migrant backgrounds. It also showed how loneliness differences are explained by different mechanisms.

Loneliness has been defined as the discrepancy between desired and actual social relationships, both in quantity and quality (Cacioppo et al., 2015). For older adults, loneliness has been found to be associated with a range of poor physical and mental health outcomes and higher all-cause mortality (Cacioppo et al., 2006; Luo et al., 2012; Rico-Uribe et al., 2018). While loneliness has traditionally been explained focusing on differences across different individual factors, researchers have recently begun to pay attention as to how the social environment, at the community and neighborhood levels, may also play an important role in contributing to loneliness (Finlay & Kobayashi, 2018; Kowitt et al., 2020), as environmental conditions could either promote or deter interactions and participation.

The concept of “neighborhood effects” has a long tradition within the social sciences (Sampson, 2011; Wilson, 1987). Neighborhood effects literature focuses predominantly on how disadvantaged neighborhoods may produce disadvantaged outcomes for individuals, a cycle that can repeat throughout generations reproducing inequality (Sharkey, 2008). It has been well documented throughout existing studies that neighborhoods, as geographic and social spaces, can also be sites of concentrated disadvantage (Ludwig et al., 2013). In his work, Sampson (2012) also suggests that there are four mechanisms that underpin neighborhood effects: social networks, social norms, infrastructure, and spatial organization. These factors could shape the ease of interactions for older adults, as well as norms around the neighborhood which can affect whether residents can have positive relationships with one another.

At the same time, due to residential segregation, we know that individuals reside in different types of neighborhoods. Particularly, racial residential segregation has been well documented in the American context (Charles, 2003; Reardon et al., 2009) as well as in other English-speaking countries (Johnston et al., 2007). Therefore, one potential pathway in which older adults from ethnic minority or migrant backgrounds may report higher levels of loneliness could be due to the conditions of their neighborhoods.

This article aims to contribute to the existing literature in three ways. First, while studies have begun to tease out whether ethnicity and migrant status are salient predictors of loneliness, current knowledge has predominantly come from countries in North America or Europe (Visser & El Fakiri, 2016; Wu & Penning, 2015). Australia is a multicultural nation that has been populated through a process of international migration (Biddle et al., 2015). The categorization of racial/ethnic/migrant status also differs from that of other countries and is defined, at a high level, by whether an individual is born in Australia or overseas, and whether someone is of an Aboriginal or Torres Strait Islander background. Within the categorization of migrants, individuals are then further defined as being from an English-speaking country or a non-English-speaking country (Biddle et al., 2015). Note that the top English-speaking migrant-sending countries are the United Kingdom and New Zealand, while non-English-speaking migrant-sending countries are predominantly from Asia, such as India, China, Philippines, and Vietnam, followed by migrants from European countries, such as Italy (Australian Bureau of Statistics, 2021). Given the specific country-level context of Australia, this article begins to build an evidence base to establish whether older adults from a migrant or Indigenous background report variation in loneliness, as compared with the majority native-born, non-Indigenous group.

Second, while studies have shown that neighborhood factors might play a role in explaining loneliness, the findings are equivocal, highlighting the need to understand how the neighborhood as a social environment might shape loneliness. Country-level differences in levels of urbanization, neighborhood segregation, and amenities within the neighborhood also render a study of Australia relevant, as its history of city and town planning necessarily differs from that of other countries. Understanding whether and how different facets of the neighborhood relate to loneliness could help inform the design of these settings or point to particular factors that might be especially important for mitigating loneliness. The majority of older Australians remain living in the community and express a preference to “aging in place” (Australian Institute of Health and Welfare, 2013), and thus, the relationship between neighborhood-level characteristics and risk of loneliness merits attention.

Third, this article investigates whether and how variation in the neighborhood environment could explain any observed differences in loneliness across ethnic–migrant groups. This may be due to residential segregation, variation in neighborhood resources across groups, or differences in perceived safety or exposure to crime that may shape social cohesion and participation in the community, with implications for loneliness. While most existing studies on loneliness are focused on individual-level factors, moving the level of focus to the characteristics of the neighborhood can show how the social environment may create more or fewer opportunities for different groups of older adults to engage in social interactions and connections. This opens up an avenue for intervention to consider a policy that can support healthy aging and can reduce disparities in health and well-being (Burgard et al., 2021).

Ethnic–Migrant Status and Variation in Loneliness

Among older adults, existing research has shown that loneliness varies by a number of individual characteristics. These include partnership status and quality of spousal relationship (Leitsch & Shiovitz-Ezra, 2010), socioeconomic status and educational attainment (Jivraj et al., 2016; Shankar et al., 2013; Smith & Victor, 2019), and available and perceived social support (Litwin & Shiovitz-Ezra, 2011). Though receiving less attention, ethnicity has also been found to be associated with loneliness, and this association has been theorized to have several possible causes.

Loneliness has been found to be more common among members of ethnic minority groups than among members of the native or dominant group (Visser & El Fakiri, 2016). Ethnic minorities tend to have fewer socioeconomic resources, due to a lower social status and psychosocial factors including experiences of discrimination and lack of social support (Stephens et al., 2011). A study comparing loneliness among Dutch, Moroccan, Turkish, and Surinamese adults in the Netherlands found that the higher levels of loneliness experienced by the ethnic minority groups were entirely mediated by ethnic differences in demographic, socioeconomic, health, neighborhood, and perceived discrimination domains (Visser & El Fakiri, 2016). They further suggested that when controlling for these domains, the ethnic minorities emerged less lonely on average than the native Dutch, which may be as a result of cultural differences. Similarly, Stephens et al. (2011) found in a study of 55- to 70-year olds in New Zealand that measures of socioeconomic status fully mediated the effects of ethnicity on health. Independent of demographic, socioeconomic, and health factors, immigration itself may also be a risk factor. In a nationally representative Canadian sample, first- and second-generation immigrants (aged 60–79 years) were lonelier than native-born Canadians or immigrants who were third generation or above. For first-generation immigrants, those who immigrated later in their life course experience greater loneliness than those who immigrated at a younger age (Wu & Penning, 2015). These findings suggest that immigrants may have more difficulty developing networks of social support, particularly if they migrated at an older age. The study also supported the broader association between loneliness and ethnic minority status, finding that non-British, non-French European immigrants were lonelier (Wu & Penning, 2015).

Indigenous peoples as an ethnic minority have specific experiences which should be considered separately from other ethnic minorities. Indigenous peoples’ suffering as a result of colonialism, institutional racism, and historical trauma poses an ongoing burden upon their mental and physical well-being (Le Grande et al., 2017). Stephens et al. (2011) outline several factors experienced by the Maori population in New Zealand that expose them to a greater risk of poor mental health, including racial discrimination, inadequate access to primary health care, differential patient management protocols, socioeconomic deprivation, and lifestyle factors. Building on prior research, the article first extends these findings to hypothesize that loneliness may also vary by ethnicity and Indigenous status in Australia among older adults.

Hypothesis 1: In the Australian context, older adults who are migrants or from an Indigenous background would report higher levels of loneliness, as compared with native-born, non-Indigenous Australians.

Neighborhood Characteristics and Loneliness

Environmental factors, such as characteristics of the neighborhood, can play a role in the experience of loneliness for older adults (Scharf & de Jong Gierveld, 2008). For example, Alidoust and Bosman (2015) outlined the key characteristics of neighborhoods relevant to social interaction and these include safety, density, walkability, and accessibility. Neighborhood characteristics, however, can be broken down into three separate components that may be relevant for loneliness: (a) neighborhood conditions that produce the context for social interactions, (b) neighborhood norms and sociality, and (c) individual satisfaction with various aspects of the neighborhood.

First, neighborhood conditions provide the context in which social interactions play out. As such, we would expect that neighborhoods that are in better conditions would allow for ease of interactions. The concept of “environmental press” is relevant here as it underscores that the physical and social environment can have psychological impacts, which lead to modified behavior (van der Greft & Fortuijn, 2017). Environmental press refers to an increase in the difficulty of navigating environments, such as the local neighborhood. Levels of crime, for instance, can be a stressor that makes the neighborhood more demanding to navigate. Older adults may respond by modifying their routines to avoid leaving the home, resulting in fewer social interactions with effects on loneliness. Empirically, one study comparing English neighborhoods combined perceived neighborhood quality with location and area socioeconomic status to create an overall measure of residential context, which the authors found to be associated with significant variations in loneliness for older adults (Scharf & de Jong Gierveld, 2008). The authors suggest that factors like housing conditions, population composition, amenities, rate of population turnover, social problems (like crime), or local policies contribute to these area differences, by affecting residents’ feelings of safety and inclination toward social participation within the neighborhood. A study of US rural-dwelling older adults also found that better-perceived neighborhood quality was associated with lower levels of loneliness (Kowitt et al., 2020). Their measure of perceived neighborhood quality encompassed the characteristics of safety, social cohesion, and neighborhood resources for physical activity and walking.

Holding neighborhood conditions constant, neighborhoods may nevertheless also vary in their levels of sociality. This may be attributed to particular social norms in the neighborhood, such as with neighbors helping out, or checking in to support one another. Neighborhood sociality may also include levels of mutual trust or mistrust among neighbors which could shape loneliness among its residents. Higher levels of sociality in the surrounding environment could also suggest that residents would identify more strongly with their neighborhood, and one study finds that increased neighborhood identification is associated with reduced loneliness (Fong et al., 2021). Finally, beyond the conditions and sociality of the neighborhood, it may be individuals’ levels of satisfaction with different aspects of their immediate environment that is related to their levels of loneliness. This captures that even within the same neighborhood, older adults may vary in their levels of satisfaction with embeddedness and connectedness with their environment. Therefore, it may be individuals’ subjective assessments of how satisfied they are with their environment that is important, independent of the conditions of the environment.

Hypothesis 2: More favorable neighborhood characteristics, including better neighborhood conditions, higher neighborhood sociality, and higher levels of satisfaction with aspects of the neighborhood, are associated with lower levels of loneliness.

Neighborhood Differences as Explaining Ethnic–Migrant Differences in Loneliness?

The association between ethnic minority status and loneliness may entail a spatial element, in at least one way, through ethnic residential segregation. Residential segregation has individual and structural elements, with residents both selecting into and out of neighborhoods, based on preferences and restrictions in local housing markets such as in the kinds of available housing stock (Hedman & Van Ham, 2012). Bécares et al. (2013) highlight that ethnic residential segregation is prevalent, and particularly for Indigenous groups, is founded in the “social manifestation of individual prejudices and institutional discrimination,” including colonial policies driving rural to urban migration and resettlement, ongoing deprivation, and discrimination in housing markets.

On the one hand, ethnic minorities derive some benefits from living in ethnically segregated areas, as a result of an “ethnic density effect” found in US, UK, and NZ studies (Bécares et al., 2013). The ethnic density effect buffers ethnic minority individuals from the consequences of discrimination and racial harassment through social cohesion, mutual social support, and a strong sense of community.

For example, a number of US studies have found that older Latinos have better health outcomes when living in neighborhoods with a higher percentage of Latinos or immigrants, and that majority-Hispanic neighborhoods are associated with better cognitive function for all residents than neighborhoods with other ethnic compositions (Aneshensel et al., 2015; Kovalchik et al., 2015). A New Zealand study of Maori adults also found that living in an area with a higher Maori concentration is associated with better self-rated health, a decreased incidence of doctor-diagnosed mental health disorders, and decreased reports of racial discrimination, personal attack, or experienced unfair treatment (Bécares et al., 2013). This suggests that immigrant enclaves may have specific social and cultural resources or practices with beneficial effects for residents, such as multilingualism, social networks, and higher levels of social integration, which may explain their continued existence (Kovalchik et al., 2015).

On the other hand, however, in their NZ study, Bécares et al. (2013) found that for Maoris, the protective effect of ethnic density was masked by area deprivation, which was independently associated with poor health outcomes and racial discrimination. The association of ethnic minority status with fewer socioeconomic resources leads ethnic minorities to be overrepresented in poor neighborhoods, which have less of the neighborhood infrastructure and higher levels of social problems, thus giving them fewer opportunities for social interactions as outlined above and potentially exposing ethnic minorities to a higher risk of loneliness.

Hypothesis 3: Given the area deprivation associated with segregated neighborhoods, variation in satisfaction with the neighborhood environment and neighborhood characteristics will mediate loneliness differences by ethnic–migrant status.

Data/Method

This study draws on four waves of data from the Household, Income, and Labour Dynamics in Australia (HILDA) Survey, with requisite measures. These were captured in Waves 6, 10, 14, and 18, in Years 2006, 2010, 2014, and 2018. The HILDA Survey is a continuous, ongoing annual longitudinal study that began in 2001 with a sampling frame of Australian households during its first year. Surveys were collected from individuals aged 15 and older living in the same households. Data collection for the HILDA Survey combines a self-complete questionnaire and computer-assisted face-to-face interviews.

The survey captures a range of information on social and economic well-being, as well as on measures of labor market and family dynamics. It also contains responses to a question on loneliness and a number of questions pertaining to the respondent’s satisfaction across different life domains and subjective assessments of their neighborhood, allowing for understanding whether and how these factors are associated with loneliness. For more information on the data set, including a detailed description of the sample design, please see Watson and Wooden (2007). Response rates for the study are generally very high, and among those who complete the main interview, approximately 90% also return a self-completion questionnaire (Watson & Wooden, 2015), which contains some of the measures used in this article.

Across the four waves of data, 81,658 observations (or person-waves) were available. As the focus of the study is on older adults, observations capturing respondents younger than age 50 were excluded, dropping 56,168 cases (N = 25,490). A further 3,307 cases were dropped, as these observations were missing on the key dependent variable, of the question on loneliness (N = 22,183 person-waves).

Missingness on the remaining variables was minimal and derives from questions around satisfaction with different life domains, neighborhood characteristics, and self-rated health, at around 1%–2% or less across different variables (Supplementary Appendix A). The only exceptions were around responses to three questions on neighborhood characteristics, which were missing between 7% and 9%. To accommodate missing data, I used multiple imputation by chained equations with m = 20 imputed data sets (Royston & White, 2011). Missing data on satisfaction with life domains, neighborhood characteristics, and self-rated health were imputed based on respondents’ reported characteristics (gender, age, relationship status, ethnic–migrant background, income, an index of residential area-level socioeconomic advantage, and educational attainment).

The final analytic sample comprised 9,305 respondents and 22,183 person-waves.

Measures

Dependent variable

The dependent variable is a question on loneliness. Loneliness comes from a single-item question within a scale on social support. The specific question asks respondents how much they agree with the statement “I often feel very lonely.” Possible responses ranged from 1 to 7, with 1 = strongly disagree and 7 = strong agree. While loneliness has been captured differently across different studies, one recent article showed that single-item measures of loneliness contribute meaningful information on the concept and are well equipped to measure loneliness (Newmyer et al., 2020). This variable has a skewed distribution and therefore, I use a log transformation of the outcome variable for analyses.

Independent variable

The main independent variable is ethnic–migrant status. Respondents are categorized into one of four possible categories: (a) Australian born, not Indigenous; (b) Australian born, Indigenous Aboriginal, and Torres Strait Islander origin (or Indigenous Australians); (c) migrant, from an English-speaking country; and (d) migrant, from a non-English-speaking country.

Mechanisms

There are two proposed mechanisms that explain ethnic/migrant differences in loneliness: satisfaction with safety, and satisfaction with the specific neighborhood and area in which individuals reside and their characteristics.

Safety satisfaction comes from responses to a set of questions around satisfaction with different life domains. The specific statement states: “I am now going to ask you some questions about how satisfied or dissatisfied you are with some of the things happening in your life. I am going to read out a list of different aspects of life and I want you to pick a number between 0 and 10 that indicates your level of satisfaction with each. The more satisfied you are, the higher the number you should pick. The less satisfied you are, the lower the number. … How safe you feel.” Possible responses range from 0 to 10, with 0 being “totally dissatisfied” and 10 being “totally satisfied.”

Neighborhood satisfaction comes from responses to the rating of satisfaction in regards to two statements: “Feeling part of your local community” and “The neighborhood in which you live.” Possible responses range from 0 to 10, with 0 being “totally dissatisfied” and 10 being “totally satisfied.”

Neighborhood sociality scale comes from a set of questions which asks respondents to evaluate the sociality of their neighborhood. This is from a question which asks “To what extent do you agree or disagree with the following statements about your neighborhood?” (1 = strongly disagree; 7 = strongly agree). This is followed with five statements, including: “This is a close-knit neighborhood,” “People in this neighborhood generally do not get along with each other,” “People in this neighborhood generally do not share the same values,” “People in this neighborhood can be trusted,” and “People around here are willing to help their neighbors.” Items were recoded when necessary such that higher values equal greater sociality. The Cronbach’s alpha for these five variables is 0.7636, indicating high internal consistency.

Neighborhood conditions scale comes from responses to the second set of questions. The questions ask respondents to rate “How common are the following things in your local neighborhood?” (1 = never happens; 5 = very common). This is followed by nine statements, including: “Noise from airplanes, trains or industry,” “Burglary and theft,” “People being hostile and aggressive,” “Homes and gardens in bad condition,” “Neighbors doing things together,” “Neighbors helping each other out,” “Rubbish and litter lying around,” “Traffic noise,” and “Vandalism and deliberate damage to property.” Items were recoded when necessary such that higher values equal better neighborhood conditions. The Cronbach’s alpha for these five variables is 0.7392, indicating high internal consistency.

Control variables

Control variables include gender, age, relationship status, income, educational attainment, an index of residential area-level socioeconomic advantage or disadvantage, and self-rated health. Respondents are categorized into one of four relationship statuses: (a) married; (b) in a de-facto relationship; (c) divorced, separated, or widowed; and (d) single. Possible categories for highest educational attainment include (a) less than Year 12, (b) Year 12, (c) vocational education or degree, and (d) college degree or higher. An index of area-level socioeconomic advantage or disadvantage (SEIFA) is a measure from the Australian Bureau of Statistics that captures the level of socioeconomic advantage or disadvantage in which the respondents live. This takes into account variables such as the proportion of families with high incomes, people with a tertiary education, and people employed in a skilled occupation. Respondents are categorized into deciles, from (1) lowest decile (most disadvantaged) to (10) highest decile (most advantaged).

Statistical Analysis

To examine the associations between ethnic–migrant status, neighborhood characteristics, and loneliness, I estimate random-effects panel regression models. These models account for the panel structure of the HILDA Survey data and are able to adjust for the nesting of observations within the same individuals up to the four survey waves (Wooldridge, 2010). To assess whether ethnic–migrant status is a predictor of loneliness, I first include a model whereby this variable is the only predictor in the model. Next, to examine whether neighborhood characteristics and satisfaction are associated with loneliness, I add in the neighborhood variables while taking out the ethnicity variable. Finally, I include a model with the ethnic–migrant status variable, the measures of neighborhood characteristics, as well as a number of control variables, to examine whether the association ethnic–migrant status and loneliness changes after the inclusion of the neighborhood characteristics measures. Analyses were conducted in STATA using xtreg.

I further tested mediation effects by using the R package “mediation” (Tingley et al., 2014) for causal mediation analysis (Imai et al., 2010). This estimates the natural direct and natural indirect effects between an exposure and outcome while adjusting for covariates (Pearl, 2012). As panel data are used, multilevel models were used to calculate the estimates. Ethnic–migrant status is treated as the main exposure variable with each group compared with the non-Indigenous, Australia-born group. The models estimate the direct effects of each ethnic–migrant status on loneliness, as well as the indirect effects of each ethnic–migrant status on each of the mediators, and each of the mediators on loneliness. The total effect is then calculated by the sum of the direct and indirect effects, and we can assess the proportion of the association that was mediated. The full mediation model includes adjustment for age, gender, relationship status, educational attainment, income, and SEIFA.

Results

Descriptive Summaries

Descriptive statistics for the sample are given in Table 1, whereby means and percentages are presented for the whole sample first and then separately by each ethnic–migrant group. Australia-born, non-Indigenous respondents comprised 68% of the cases, while migrants from English-speaking countries and non-English-speaking countries comprised 15% and 16% of the observations, respectively. The remaining 2% of the cases were from Australia-born, Indigenous respondents of Aboriginal or Torres Strait Islander origin. I performed analysis of variance tests to determine differences by ethnic–migrant statuses, but find across all variables that differences were significant. As post hoc tests, I performed the Scheffe’s test to compare which pairs of means were significantly different at the 0.05 level (Salkind, 2010). These are presented in the table.

In terms of the outcome variable, of loneliness, migrants from non-English-speaking countries and Indigenous Australians report the highest levels of loneliness, at 2.99 (on a scale from 1 to 7) and 3.03, respectively. Migrants from English-speaking countries report the lowest levels of loneliness, at 2.66, and this is even lower than those Australia-born from a non-Indigenous background, at 2.81.

In terms of reports of characteristics of their neighborhoods, Australia-born, non-Indigenous respondents rate their neighborhoods highest in sociality (4.68), while Indigenous Australians rate theirs the lowest (4.31). Beyond the sociality of the neighborhoods, however, Australia-born, non-Indigenous respondents and migrants from English-speaking countries rate their neighborhoods as having the best conditions (3.47), while Indigenous Australians rate their neighborhoods as having worse conditions (3.26).

Turning to sociodemographic characteristics, migrants from English-speaking countries report the highest scores on self-rated health, while Australia-born, Indigenous respondents report the poorest health. In terms of relationship status, Indigenous respondents are overrepresented in terms of being single (at 15%) and underrepresented in terms of being married (at 35%). For educational attainment, migrants from non-English-speaking countries are overrepresented in terms of having a college degree or higher (at 28%), while Australia-born respondents of Aboriginal and Torres Strait Islander origin are overrepresented in terms of having less than a Year 12 education (at 49%). For income, migrants from English-speaking countries report the highest mean income, at AUD$37,433, while Indigenous Australians report the lowest income, at AUD$27,278. Migrants from English- and non-English-speaking countries live in areas that are relatively more advantaged, while Indigenous Australians live in areas that are the most socioeconomically disadvantaged.

Results

Table 2 reports findings for differences in loneliness by ethnic–migrant status. Model 1 shows that migrants from a non-English-speaking country report higher levels of loneliness (0.069; p < .01), as compared with Australia-born, non-Indigenous respondents. Model 2 begins to explore the role of neighborhood characteristics. Across the models, it shows that more favorable neighborhood characteristics are consistently related to lower levels of loneliness, across neighborhood conditions (−0.042; p < .001), neighborhood sociality (−0.051; p < .001), subjective satisfaction with safety (−0.042; p < .001), feeling part of one’s local community (−0.026; p < .001), and satisfaction with one’s neighborhood (−0.012; p < .001).

Model 3 demonstrates whether ethnic−migrant differences in loneliness remain after the inclusion of measures capturing neighborhood characteristics and a number of covariates. As shown, the coefficient for migrants from a non-English-speaking country has attenuated (0.042; p < .05); nevertheless, it remains significant. The associations between various neighborhood characteristics and loneliness remain robust. Turning to demographic characteristics, each unit increase in age for this sample is associated with lower levels of loneliness (−0.002; p < .001). Respondents with vocational education or degree (0.032; p < .05), as well as with less than Year 12 education, report higher loneliness (0.063; p < .001), as compared with those with a college degree or higher. Compared with respondents who are married, those in a de-facto relationship report higher levels of loneliness (0.047; p < .05), as do those who are divorced/separated/widowed (0.247; p < .001), and those who are single (0.202; p < .001). Consistent with prior research, higher income is also associated with lower levels of loneliness (−0.000; p < .01). Beyond individual-level characteristics, the model also shows that respondents who live in areas that are more socioeconomically advantaged report lower levels of loneliness (−0.007; p < .001).

Supplementary Appendix B presents results from the causal mediation models as a formal test of mediation. Across all models, there were significant differences between Australia-born, non-Indigenous older adults and older adults who were migrants from a non-English-speaking country. This can be observed in the significant total and direct effects for the mediators (see Models 1–5). The results suggest little difference in levels of loneliness between Australian-born and migrants from English-speaking countries, however. We also do not observe any total or direct effects for Aboriginal or Torres Strait Islander older adults. We caution that this may be due to an issue of statistical power, as there are few cases of Aboriginal or Torres Strait Islander older adults.

Again, results show that when comparing older adults who are migrants from non-English-speaking countries (NESB) with the Australia-born, non-Indigenous respondents (AB), significant total effects can be seen for all five examined mediators. Indirect effects are significant for all of the mediators except for neighborhood conditions. More specifically, neighborhood sociality mediated 10.8% (p < .001) of the total effect or differences in loneliness between the two groups. Satisfaction with safety mediated 9.5% (p < .001) of the total effect. Satisfaction with feeling part of community mediated 8.1% (p < .001) of the total effect. And satisfaction with neighborhood where respondents live mediated 7.8% (p < .001) of the total effect. Generally, the results indicate that neighborhood sociality and neighborhood satisfaction are significant mediators of the reported difference in loneliness between older adults who are migrants from non-English-speaking countries (NESB) with the non-Indigenous, Australia-born (AB).

Discussion

Loneliness has been the focus of much attention in recent years, with the appointment of Ministers of Loneliness in countries such as the United Kingdom and Japan. Concerns about loneliness have also been highlighted given the distancing measures and lockdown during the coronavirus disease 2019 (COVID-19) pandemic. While the predominant focus of research has been on using individual characteristics to explain variation in loneliness, researchers increasingly have considered the social environment in which individuals are embedded, including their neighborhood and community, as additional factors to help understand variation in loneliness. This is particularly salient as the social environment may promote or hinder social interactions and engagement for older adults.

While studies from North America and Europe have reported variation in loneliness among older adults of different ethnic and migrant backgrounds, differences in the intake of migrants from different parts of the world, the incorporation and integration of migrants, and variation in the rhetoric around the role of immigrants in general necessarily shape their experiences and social participation and engagement in the community. To the author’s knowledge, this is the first study that has examined potential differences in loneliness among older adults in the Australian context by ethnicity and migrant status and therefore contributes to the evidence base for understanding ethnic–migrant differences in the aging experience.

In line with prior studies, this article finds that migrants from non-English-speaking countries report higher levels of loneliness, as compared with non-Indigenous, native-born Australians. The finding of migrants from English-speaking countries reporting similar experiences as the non-Indigenous, native-born is an important one; however, as it suggests that rather than migrant status per se driving loneliness differences, it could be through either (a) language barriers and/or (b) the race of the respondents that shape their daily interactions with others. Future studies that examine how these factors play a role in the aging experience of older adults who are migrants from non-English-speaking countries could be important and point to interventions that could improve their well-being.

Findings from this study also point to future directions of research in terms of highlighting divergences in the conditions of the neighborhood and perceptions and satisfaction as reported by respondents as worthy of further investigation. Noteworthy is that while migrants from non-English-speaking countries reside in areas that are relatively more socioeconomically advantaged than that of the native-born (Table 1), and also report lower prevalence of burglary and theft, and vandalism and deliberate damage to property in their neighborhoods, this does not necessarily translate into higher levels of safety satisfaction. In contrast, while Indigenous Australians report living in the most socioeconomically disadvantaged areas, they report comparatively high levels of safety satisfaction and also satisfaction with feeling part of the local community. This underscores that future research that captures more direct measures of social interactions, experiences, and behaviors of respondents would be valuable. It also suggests the need to better understand the meaning of safety satisfaction for migrants from non-English-speaking countries.

Though this study has contributed to existing knowledge on loneliness, it is not without limitations. First, the latest wave of data used in the analyses was from Year 2018, prior to the start of the COVID-19 pandemic. Given its societal impacts, future research drawing on data collected since 2020 would be useful. For ethnic minorities and migrants, recent events in 2020, such as the Black Lives Matter movement and the rise in physical violence against people of Asian descent in the United States highlighted how the everyday lived experiences of racial and ethnic minorities and migrants might differ from that of individuals from majority groups (Human Rights Watch, 2020; Lee, 2021; Pew Research, 2020; Wu et al., 2021). Such social trends in the United States are also observed in Australia. A study conducted in October 2020 by the Australian National University found 84.5% of Asian Australians reported one form of discrimination, such as at the workplace, when renting or buying a house, or at a shop or restaurant (Biddle et al., 2020). This compares with 38% of non-Asian Australians reporting having experienced any such form of discrimination. A study published in 2021 by the Lowry Institute, an independent think tank, also found that 18% of Chinese Australians reported being physically threatened or attacked in the past year (Lowry Institute, 2021). At the same time, racism experienced by Indigenous Australians has been well documented (Markwick et al., 2019; Paradies & Cunningham, 2009). Experiences in public spaces such as in and around one’s neighborhood might therefore be of concern and a lens through which to understand differences in reports of loneliness for minority older adults as compared with the ethnic majority. Second, measures of neighborhood characteristics are subjective assessments and may be open to bias. Research considering objective indicators, in addition to subjective reports of neighborhoods may also be helpful. Despite these limitations, the rich data from the survey provide new insights into whether and how migrants and Indigenous Australians may report higher levels of loneliness. This article is also able to point to future research that can capture different aging experiences of older adults, as well as how ethnicity and migrant status are important dimensions that shape this process. Given the Australian population is aging, while at the same time becoming more diverse, more attention is needed to understand variability in health and well-being outcomes by ethnicity and migrant status, as well as better integration of the social context into the study of such factors leading to health and well-being disparities.

Funding

This research was partially supported by the Australian Research Council Centre of Excellence for Children and Families over the Life Course (project number CE200100025) and by an Australian Research Council Discovery Early Career Researcher Award (project number DE210100582). The Center is administered by the Institute for Social Science Research at the University of Queensland, with nodes at the University of Western Australia, the University of Melbourne, and the University of Sydney. This article uses unit record data from the Household, Income, and Labour Dynamics in Australia (HILDA) Survey. The HILDA Project was initiated and is funded by the Australian Government Department of Social Services (DSS) and is managed by the Melbourne Institute of Applied Economic and Social Research (Melbourne Institute). The findings and views reported in this article, however, are those of the author and should not be attributed to either DSS or the Melbourne Institute.

Conflict of Interest

None declared.

Acknowledgments

The author would also like to thank the helpful feedback of the reviewers and Michael Vuolo and Ryan Moltz. The author would also like to thank the research assistance of Yongbo Liu.

References