Abstract
Age at arrival is one factor that could influence the integration of humanitarian migrants, especially for children and teenagers. Previous research has focused on the influence of the age at arrival on education, employment, social and language learning outcomes, but there is limited research, especially for longitudinal study, on other important measures of integration. Moreover, young adult and adolescent refugees, and the relationship between age and integration outcomes are under-studied. To address these gaps, this study examined the relationships between age at arrival and different dimensions of integration of young refugees in Australia, using five years’ panel data from the Building a New Life in Australia (BNLA) longitudinal study from 2013 to 2018 (282 individuals). Our findings indicate that age at arrival is significantly correlated with multiple integration outcomes. Old entrants tend to have a higher probability of having a paid job but have poorer mental health and English proficiency. On the other hand, older entrants were more likely to know their rights well compared to younger entrants at arrival. However, younger entrants’ knowledge of their rights overtook older entrants four to five years after their arrival. These findings suggest that a shorter assessment process and enabling earlier entry especially among refugee youth and young adults could effectively improve their future settlement outcomes under the current humanitarian policies and system.
Key messages
Age at arrival is significantly related to the integration outcomes of adolescents and young adults, and therefore enabling them to enter at an earlier age could improve their f uture outcomes.
This study examines the correlations between age and integration, and how they may change over time.
The first longitudinal study focuses on age and integration of refugees in the Australian context.
Introduction
Refugee integration is a complex and multidimensional process of immigrants becoming fully fledged members of the host communities, which comprises factors such as education, employment, and social participation (Ager and Strang, 2008). Age at arrival is one factor that could influence the integration of migrants, especially for children and teenagers (Gonzales, 2003; Myers et al, 2009; Bleakley and Chin, 2010; Åslund et al, 2015; Hermansen, 2017). Previous research has focused on the influence of the age at arrival on education, employment, social and language learning outcomes, but there is limited research on other important measures of integration. Most integration studies have been based on cross-sectional data and fail to examine integration outcomes over time. Moreover, young adult and adolescent refugees, and the relationship between age and integration outcomes are under-studied (Pritchard et al, 2019). To address these gaps, this study seeks to understand the relationships between age at arrival and different dimensions of integration of young refugees in Australia, using five years of data from the Building a New Life in Australia (BNLA) longitudinal study.
Age at arrival and biological age
There have been many studies conducted on the effects of the age at immigration on young immigrants’ future outcomes, as the timing of immigration could influence the timing of life course events and transitions and hence the integration trajectories of immigrants. The timing of immigration could delay the beginning of school and desired years of education (Correa-Velez et al, 2016). Parents could also treat their children differently and have different sets of expectations due to socio-cultural norms and expectations depending on their age (Bornstein, 2012). A delay in the timing of specific life course events, such as language acquisition and academic performance, could influence the future outcomes of individuals as there are sensitive periods for learning and skill development (Myers et al, 2009; Bleakley and Chin, 2010; Basu, 2018; Lemmermann and Riphahn, 2018). Therefore, age at arrival could have a substantial effect on the integration outcomes of immigrants, whereas most research on the effect of the age at arrival on integration and well-being focuses on immigrants in general and not refugees.
Age at arrival and education
A large body of research has indicated that age at arrival could influence the educational outcomes of immigrants and refugees, with the age at arrival negatively correlated with the years of schooling, academic performance, educational attainment and college/high-school drop-out rates (Gonzales, 2003; Cortes, 2006; van Ours and Veenman, 2006; Böhlmark, 2008; Cohen Goldner and Epstein, 2014; Åslund et al, 2015; Correa-Velez et al, 2016; Hermansen, 2017; Basu, 2018; Lemmermann and Riphahn, 2018). For example, Basu (2018) used sibling fixed effects to examine age at arrival on immigrants’ education and reported that teenage immigrants had worse education outcomes compared to child immigrants. Correa-Velez et al (2016) conducted a longitudinal study on refugee’s education outcomes after eight to nine years in Melbourne, Australia, and found that the younger the age of arrival, the higher the odds of secondary education completion. However, contrary to other studies, language proficiency did not predict school completion and discrimination was a significant contributor to secondary school drop-out (Correa-Velez et al, 2016).
Some research also suggests that the effect of age at arrival on education may not be significant if other factors such as language proficiency and citizenship are taken into account (Beck et al, 2012; Basu, 2018). However, Lemmermann and Riphahn (2018) argue that although language skill is a key factor in educational outcomes, language skills alone cannot fully explain the link between age at migration and educational attainment. Most of the studies to date have focused on the effect of age of arrival for immigrant children and youth but not specifically for refugees.
Age at arrival and language skills
Research has shown that immigrants who arrived at an earlier age are more likely to report higher levels of English proficiency (Bleakley and Chin, 2010; Beck et al, 2012; Basu, 2018). Beck and colleagues (2012) examined the effect of age at immigration on adult attainment using the 2000 US census. They reported that the chances of reporting a high level of proficiency in English fell by about 1.7 percentage points per year for each extra year of age at arrival beyond age eight (Beck et al, 2012). Similar research conducted by Bleakley and Chin (2010) also found that for each extra year of age at arrival past age nine, the probabilities of ‘speaking any English’, ‘speaking English well’ and ‘speaking English very well’ decreased by 0.6, 3 and 7 percentage points respectively.
Sensitive periods for language and education
Sensitive periods have been identified for language acquisition and education outcomes in immigrants. A sensitive period is a period of limited time when a child would be more capable of developing certain skills, such as language acquisition, than at other age. Many studies have shown that there is a significant impact on language acquisition and education outcomes when arriving after pre-school age and after primary school age (Gonzales, 2003; Böhlmark, 2008; Beck et al, 2012; Cohen Goldner and Epstein, 2014; Basu, 2018; Lemmermann and Riphahn, 2018). Different sensitive periods are observed when using different samples and in different contexts. According to Lemmermann and Riphahn (2018), sensitive periods can be as early as six (learning declines significantly after six years old), other studies suggest that the sensitive period starts at about eight to nine years old (Gonzalez, 2003; Böhlmark, 2008; Beck et al, 2012; Basu, 2018) and in some studies of immigrants there can be two sensitive periods prior to adolescence (Myers et al, 2009). Research in Israel also suggests that the effect of age at arrival on education varies with the country of origin and country of resettlement, as the education system, socio-cultural settings and life course experiences of children and teenagers are different in different countries (Cohen Goldner and Epstein, 2014).
Age at arrival and other integration outcomes
Empirical research also indicates that age at arrival could influence future employment, health and social outcomes. Immigrants tend to have better employment outcomes if they arrived at younger ages (Gonzales, 2003; Correa-Velez et al, 2016; Hermansen, 2017). Some evidence has shown that the future earnings and socio-economic status of immigrants are negatively associated with the age at arrival (Gonzalez, 2003; Correa-Velez et al, 2016). However, age at arrival may influence employment outcomes indirectly through improved language proficiency and education (Hermansen, 2017). Hermansen (2017) examined the age at arrival and life chances among child immigrants, using panel data from the Norwegian administrative registries. Results from the sibling fixed effect models suggested that older entrants to Norway had lower educational attainment, less earnings and a lower probability of having a stable job, but a higher chance of being on social welfare (Hermansen, 2017). Although education mediated the effect of age at arrival on employment, it did not mediate the significant effect of age at arrival on earnings, occupational outcome and the likelihood of receiving social welfare (Hermansen, 2017).
Research has shown that the age at arrival could influence the subsequent tendency of social segregation (Myers et al, 2009; Beck et al, 2012) and social assimilation (Bleakley and Chin, 2010). Immigrants who arrived at an older age were more likely to be socially segregated (Bleakley and Chin, 2010) and be less acculturated (Kuo and Roysircar, 2004). Moreover, Kuo and Roysircar (2004) also suggested that age at arrival was associated with more acculturative stress (reduction in physical and mental health as a result of acculturation).
Age at arrival can also influence the physical and mental health outcomes of refugees. Older entrants tended to associate with more severe general distress and somatic distress (Brown et al, 2010), reduced physical growth (delayed puberty and malnutrition) (Pak, 2010) and engagement with the healthcare system (Fadnes and Diaz, 2017). Moreover, in a study of 38 Afghanistan refugee youth and young adults who resettled in the US, Mghir and colleagues (1995) discovered that older entrants have higher risk of post-traumatic stress disorder (PTSD) and depression. However, the sample size is small and the results need replication.
Although empirical research has shed the light on the relationship between age at arrival and most of the aspects of integration, understanding the relationship between age at arrival and mental health and rights is limited and the majority of research focuses on specific integration outcomes, whereas integration is a complex and multidimensional process that occurs over many years.
Refugee integration: Ager and Strang’s framework
According to the United Nations High Commissioner for Refugees (UNHCR), integration is a dynamic and multifaceted process, which requires efforts by both the host communities and refugees (UNHCR Australia, 2020). The predominant framework of refugee integration is that by Ager and Strang (2008). It includes ten specific domains within four areas of attainment: markers and means (employment, housing, education, health); social connection (social bridges, social bonds, social links); facilitators (language and cultural knowledge, safety and stability); and foundation (rights and citizenship). Recently, the UK’s Home Office also introduced two new domains: leisure and digital skills to this framework (Home Office, 2019). Integration is a multidimensional and enduring process. However, to date, there has been only one longitudinal study testing this multidimensional model of integration (Lichtenstein and Puma, 2019). Findings from this four-year study suggested that refugees were more integrated over time as they stayed longer in the country of resettlement. Moreover, they discovered that integration is a phased phenomenon in that the different domains of integration are not improving at a constant and same pace (Lichtenstein and Puma, 2019). While this study highlighted the multidimensionality of integration and its development over time it did not explore changes in the integration experience by age of arrival.
Emerging adulthood, the integration of refugee youth and the current study
This study focuses on the emerging adulthood population instead of only the under-18 population, as emerging adult refugees could be more vulnerable, and disadvantaged compared to refugees from other age groups as it is a period of great transition in the life course. Emerging adulthood refers to young adults aged 18–25 (Arnett, 2000; 2004; Arnett et al, 2014). Arnett et al (2014) argue that there are many demographic changes during this period compared to other life stages and a period of preparation for adulthood. Emerging adulthood is characterised by (1) identity exploration, formation and development; (2) instability in relationships and identities; (3) self-focus as there are fewer social rules and obligations; (4) feelings of being in-between adolescence and adulthood and (5) optimism and confidence about the future (Arnett, 2000; Arnett et al, 2014). The duration of emerging adulthood varies between countries, as there are different social clocks and social expectations in different countries (Arnett et al, 2014). The integration process of refugee youth during emerging adulthood is particularly important given normative transitions and expectations in their birth country were greatly disrupted because of their refugee experiences and the adjustment to new social clocks and expectations in the host country is part of integration for this cohort.
In this study, we examined refugees in their late adolescence or emerging adulthood at arrival (14–21 years in Wave 1) and followed them for five years when some of them are transitioning from adolescence to emerging adulthood, and emerging adulthood to adulthood (18–26 years in Wave 5). To date, research on the effect of the effect of age at arrival on different dimensions of integration outcomes has been limited to a few rather than multiple integration outcomes and to cross-sectional studies. Therefore, this study aims to fill the gap in the literature by examining the effect of age at arrival on different dimensions of integration of youth and emerging adult refugees in Australia, using the five years of data from the BNLA longitudinal study.
Methods
Data and sample
Five waves of data from BNLA data were employed in this study. BNLA is a longitudinal cohort study funded by the Australian government to provide information on the settlement of humanitarian migrants who arrived in Australia in 2013. BNLA monitored their settlement outcomes, such as the housing, employment, health (both mental and physical), social and family conditions over five years annually (five waves) (Edwards et al, 2018). We use data from Waves 1, 3 and 5, which were collected face-to-face using computer-assisted personal interview or computer-assisted self-interview (Edwards et al, 2018). The interview was assisted by bilingual interviews and the interviews were conducted in English and other translated languages (Edwards et al, 2018). The wave response rates have been higher than 80% in each wave and the retention rate of the Wave 1 sample in Wave 5 was 78% (BNLA, 2019). The BNLA study was granted ethics approval from the Australian Institute of Family Studies Human Research Ethics Committee.
In Wave 1 there were 2,399 participants, who were granted a permanent humanitarian visa in the three to six months prior to commencement. There are 1,509 Primary Applicants who are named on the visa application for the family unit, 755 adult Secondary Applicants and 135 adolescent Secondary Applicants who were above 15 years old and residing with the Primary Applicant at the time of the Wave 1 data collection. Nineteen languages were used in the Wave 1 data collection (Edwards et al, 2018).
We focus on participants under 22 years old at Wave 1 that responded to the Waves 3 and 5 surveys (n = 282). At Wave 1 these participants were aged 14–21 (M = 18.1, SD = 1.97, Median = 18). Among them, 139 were male and 143 were female and the majority were born in Iraq (n = 102), followed by Afghanistan (n = 74) and Iran (n = 55). As for the pre-arrival visa subclass, the majority of the sample arrived under a 200 Refugee Visa (66%) and 204 Women at Risk Visa (22%) (Table 1).
Demographic characteristics of the sample
| Frequency (%) | |
|---|---|
| Age at arrival | |
| 14 | 4 (1%) |
| 15 | 30 (11%) |
| 16 | 40 (14%) |
| 17 | 36 (13%) |
| 18 | 42 (15%) |
| 19 | 51 (18%) |
| 20 | 39 (14%) |
| 21 | 40 (14%) |
| Country of birth | |
| Iraq | 102 (36%) |
| Afghanistan | 74 (26%) |
| Iran | 55 (20%) |
| Nepal | 18 (6%) |
| Myanmar | 13 (5%) |
| Pakistan | 6 (2%) |
| Congo (Democratic Republic) | 6 (2%) |
| Syria | 3 (1%) |
| Bhutan | 3 (1%) |
| Sri Lanka | 1 (0.4%) |
| Eritrea | 1 (0.4%) |
| Visa subclass (pre-arrival) | |
| 200 Refugee | 187 (66%) |
| 204 Women at Risk | 62 (22%) |
| Special Humanitarian Programme | 18 (6%) |
| Onshore Protection Visa | 15 (5%) |
Measures
Independent variables
Age at arrival
The exact age at arrival was confidentialised and not released in the BNLA study. Instead, it provided the information of the age at interview and the different periods of the length of stay prior to the interview at Wave 1. Therefore, age at arrival will be measured using the age at interview in Wave 1. Although there was a gap between the age at interview and the age at arrival, all the participants in the sample population arrived less than one year before the interview and 90% of the sample arrived in Australia three to five months prior to the interview and therefore, the difference between age at first interview and age at arrival was small.
Length of stay
The effect of the length of stay will also be estimated using wave as a dummy indicator variable. There is an approximately two-year gap between Waves 1 and 3, and Waves 3 and 5.
Dependent variables
Employment
In this study, participants are asked whether they have a paid job in the seven days prior to the interview.
Health
Physical health: assessed using self-reported physical health conditions. Participants rated their physical health during the 4 weeks prior to the interview. It was asked using a six-point scale, from 1 (Excellent) to 6 (very poor). In this study, we generated a dummy variable for physical health, with 0 being ‘1 = Excellent’, ‘2 = Very good’, ‘3 = Good’ and 1 being ‘4 = Fair’, ‘5 = Poor’ and ‘6 = Very poor’.
PTSD: Post-Traumatic Stress Disorder-8 (PTSD-8) scale is an eight-item measure for probable PTSD comprising four intrusion, two avoidance and two hypervigilance items assessing the likelihood of having PTSD (Hansen et al, 2010). Participants rated symptoms on a four-point Likert scale (1 = not at all to 4 = most of the time). Probable diagnosis of PTSD is met if at least one symptom from each of the three PTSD-8 subscales has an item score of 3 or 4 (Hansen et al, 2010). A dummy variable of probable clinical significant PTSD (1 = PTSD problems and 0 = not having PTSD) is used in the BNLA survey to identify refugees who might have PTSD.
PTSD-8 score is also calculated in this study as the indicator of PTSD symptoms using the PTSD-8 scale. The minimum score was 4 and the maximum score was 32. A high score indicates more severe PTSD symptoms and hence a higher risk of having PTSD.
Psychological distress (Kessler-6 Scale): K-6 score and a dummy variable of K-6 groupings (1 = probable serious mental illness and 0 = no probable mental illness) were used to identify psychological distress and probable serious mental illness. The Kessler-6 Scale comprises six items and five response categories using the Australia scoring scale (ranging from 1 = all of the time to 5 = all of the time) for each item, indexing two anxiety symptoms and four depression symptoms. The minimum score is 6 and the maximum score is 30. A score between 19 and 30 indicates having probable serious mental illness (Australian Bureau of Statistics, 2012). Thus, a higher K-6 score is associated with a higher possibility of serious mental illness.
Social bonds and social bridges
Participants are asked whether their friends in Australia are: 1 = Mostly from my ethnic/religious community; 2 = Mostly from other ethnic/religious community; 3 = A mixture; or 4 = Do not have any friends in Australia yet. To measure how well they connect with people outside their ethnic group in the host community, we coded 1 as having friends with a mixture of backgrounds or mainly from ethnic groups other than their own and 0 being not having any friend or having friends mainly from their own ethnic group.
Language and cultural understanding
A four-item scale measured English proficiency. Participants answered how well they read, speak, understand and write English, with a four-point scale, from 1 (very well) to 4 (not at all). The scale was recoded to 1 (not at all) to 4 (very well) for this study. The scores for each of the items are summed to generate a score for overall English proficiency 4 (low English proficiency) to 16 (high English proficiency).
As for the understanding of Australia culture, this was measured by a four-point scale ranging from 1 (very easy) to 4 (very hard) in the BNLA study. A dummy variable (0 = ‘easy’ and ‘very easy’; 1 = ‘hard’ and ‘very hard’) was derived.
Security and stability
Security and stability was measured by interpersonal trust and the sense of belonging.
Interpersonal trust: measured by an aggregate trust score participants have on different groups of people in the host community. Participants answered the question ‘To what extent (1 = A lot to 4 = Not at all) do the participants trust different groups of people?’ (their neighbourhood, the wider Australian community, the police, the people they work/study with, and the media). We reversed the scale and summed the score to create a trust score variable.
Sense of belonging: participants were asked to indicate how often they feel being part of the Australia community, from 1 (Never) to 5 (Always). We created a dummy variable for sense of belonging, with 1 being having at least some sense of belonging, and 0 being hardly never and never.
Rights and citizenship
Participants were asked how well they know how to find out their own rights (for example, legal rights, tenancy rights), with a four-item scale measure (1 = know very well; 2 = know fairly well; 3 = know a little; 4 = wouldn’t know at all). For analytic purposes, we constructed a dummy variable, with 1 being ‘know very well’ and ‘know fairly well’, and 0 being ‘know a little’ and ‘wouldn’t know at all’.
Covariates
Covariates are: country of birth, sex, visa subclass before arrival. The country of birth was a categorical variable with refugees who were born in Pakistan, Congo (Democratic Republic), Syria, Bhutan, Sri Lanka and Eritrea grouped as ‘other countries’ because of the small number of observations. Sex was a dummy variable containing male and female. Visa subclass arrival was a categorical variable that recorded the visa class (visa subclass 200, 202, 204, 866) of the humanitarian visa they held before their resettlement in Australia. Visas 200, 202 and 204 are offshore permanent protection visas which require refugees to apply outside of Australia, whereas visa 866 is an offshore protection visa with different branches. However, for onshore protections visas (visa subclass 866), we categorised them as Onshore Protection visas because of the small number of observations.
Analytic approach
To capture the relationships between age of arrival and refugee integration outcomes, random effects (RE) multiple linear regression models were used to analyse the repeated longitudinal measure of continuous measures (K-6 scores, PTSD-8 scores, English proficiency and interpersonal trust) and RE multiple logistic regression models were used for the binary outcomes. RE linear regression models can estimate both within-individual and between-individual effects and accounting for the correlation of the repeated outcomes in the same individual (Bosker and Snijders, 2011). For each model, the intra-class correlation (ICC) was estimated. ICC indicates the relatedness of the group data by comparing the variance between and within groups (Bosker and Snijders, 2011). RE were used instead of fixed effects because age at arrival is time invariant. There are missing data due to attrition; however, the non-response rate (<20%) is relatively low compared to other longitudinal surveys in Australia. Conditional on some covariates the missing data are likely to be missing at random in this research. We also tested for the multicollinearity of the independent variables. Aside from the interaction between age and wave, which is correlated by design, the variance inflation factor did not suggest multicollinearity was an issue for any of the variables in any of the models. We used Stata IC 15.1 for all the analyses.
Results
This section describes the descriptive and regression results for each of the outcome variables. The demographic characteristics of the analytic sample are presented in Table 1. There were 282 refugees who responded to Waves 1, 3 and 5 of the BNLA survey who were 14–21 years old at Wave 1. The largest age group was 19 years old (18%) and the smallest age group was 14 (1%) at Wave 1. Refugees were from 11 countries and the majority of them were from Iraq (36%), Afghanistan (26%) and Iran (20%). Some 66% of the sample held a 200 Refugee visa and about 22% held a 204 Women at Risk visa before arrival (for a description of the visas see Edwards et al, 2018). Descriptive results for integration are presented in the Appendix A (Table A1).
Markers and means
Mental health
For psychological distress, mean scores were 11.67, 12.03 and 11.17 in Waves 1, 3 and 5 respectively. As for the likelihood of having serious mental health problems, 11–12% of the sample were likely to have mental health problems at Waves 1, 3 and 5.
Results in Table 2 suggest that age at arrival was significantly associated with psychological distress (p < .05) but length of stay and age by wave was not. Females had higher levels of psychological distress. RE models also show that neither the age at arrival nor age by wave was significantly associated with mental health problems. Females were more likely to have mental health problems and, compared to Iranian-born refugees, Iraqi-born refugees were more likely to have mental health problems.
The association between age at arrival and length of stay (wave) on integration, random effects linear and logistic regression models
| Age at arrival | Wave | Age by wave | ||
|---|---|---|---|---|
| 3 | 5 | |||
| Psychological distress | 3.65** | 0.83 | 0.55 | −0.11 |
| Mental health problems1 | 1.06 | 0.82 | 0.55 | 1.01 |
| Post-traumatic stress disorder symptoms | 0.32 | 1.86 | 2.59 | −0.33 |
| Post-traumatic stress disorder problems1 | 1.26** | 8.62* | 57.87 | 0.96 |
| Fair or poor health1 | 0.98 | 0.07* | 0.002*** | 1.06* |
| Employed | 1.62** | 197.52*** | 14346.01** | 0.94 |
| Mixture of friends1 | 0.87* | 3.42 | 4.15 | 1.00 |
| English | −0.25*** | 2.12*** | 4.02 | −0.02 |
| Cultural understanding1 | 0.96 | 0.02*** | 0.002*** | 1.08*** |
| Trust | −0.10 | −1.03 | −1.78 | 0.02 |
| Sense of belonging1 | 1.07 | 35.57 | 225.18 | 0.97 |
| Knowledge of rights1 | 1.20** | 82.15*** | 9164.94*** | 0.93*** |
Notes:
Odds ratios.
Random effects models also included the covariates: sex, country of birth, visa subclass before arrival.
All results are based on the 14–21 years old sample, except for employment, as working under 15 years old is illegal in Australia.
Full models can be found in the Appendix.
* p < .1; ** p < .05; *** p < 01
PTSD
Between 19% and 23% were likely to have PTSD across Waves 1, 3 and 5.
Age at arrival was significantly related to the odds of having clinically significant PTSD. For each extra year of age at arrival, the odds of having PTSD increased by 26% (p < .05). Clinical levels of PTSD were significantly higher at Wave 3 compared to Wave 1 (with 90% confidence) but not at Wave 5. Figure 1 shows the estimated the relationship between age at arrival and the probability of clinical levels of PTSD. It shows that the association between age at arrival and clinical levels of PTSD was statistically significant at Waves 1 and 3 but not at Wave 5.
Probability of having clinically significant PTSD by age at arrival, Waves 1, 3 and 5
Citation: Longitudinal and Life Course Studies 2023; 10.1332/175795923X16839016657163
Probability of having clinically significant PTSD by age at arrival, Waves 1, 3 and 5
Citation: Longitudinal and Life Course Studies 2023; 10.1332/175795923X16839016657163
Probability of having clinically significant PTSD by age at arrival, Waves 1, 3 and 5
Citation: Longitudinal and Life Course Studies 2023; 10.1332/175795923X16839016657163
Contrary to the regression results of clinically significant PTSD, there was no significant relationship between age at arrival and PTSD symptoms. Females (compared to males) and Iraqi-born (compared to other countries) were more likely to have clinical levels of PTSD and PTSD symptoms.
Physical health
The percentage of refugees who had relatively poor physical health was 19%, 18% and 13%, at Waves 1, 3 and 5 respectively. Age at arrival was not significantly associated with poor physical health, whereas wave and age by wave (with 90% confidence) were correlated.
In terms of demographic correlates, the odds of having poor health was higher for females (compared to males), those holding an Onshore Protection visa, or a Women at Risk visa and being born in Nepal (compared to other countries). Compared to holders of 200 Refugee visa the odds of having poor health were about 1.8 times less likely for Women at Risk visa holder and five times less likely for Onshore Protection visa holder.
Employment
At Wave 1, 2% of the sample reported that they had paid work within the seven days prior to the interview. The proportion of participants who had paid work increased to 17% at Wave 3 and 32% at Wave 5. Age at arrival and the length of stay were significantly correlated with employment. For each extra year of age at arrival, the odds of getting a job increased by 62% (p < .05). As for the length of stay (wave), the odds for people having a paid job in Waves 3 and 5 were 198 times and 14,346 times higher than in Wave 1 respectively. The extremely large odds ratios could be explained by the small number of participants who were employed in Wave 1 (2%) and by the fact that it is only legal to work at 15 years of age in Australia.
Furthermore, sex, country of birth (Iraq and Iran) and holding an Onshore Protection visa before resettling into Australia were also significantly related to the employment. The odds of females having a job was 88% lower compared to males. In term of the country of birth, Iraqi and Iranian refugees had 70% and 82% lower odds of getting a job compared to refugees from other country of birth.
Social connection
Ethnic religious background of friends: The number of participants who made friends with ‘A mixture’ of background or people ‘Mostly from other ethnic/religious community’ were 43%, 66%, 68% at Waves 1, 3 and 5 respectively. The model shows that the odds of having friend outside their own religious/cultural background reduced by about 14% for each extra year of age at arrival (p < .1). The probability of refugees who arrived at 14 years old (70%) of having friends mostly outside their ethnic/religious group or a mixture of friends was about 19 percentage points higher than refugees who arrived at age 21 (51%), after controlling for the demographic characteristics.
Moreover, the odds of females having a mixture of friends or friends mostly outside their ethnic/religious community was 30% lower than males, whereas the odds of refugees born in Iran, Iraq, Myanmar and Afghanistan were 58%, 53%, 90% and 68% lower compared to those who are from ‘other’ country of birth respectively.
Facilitators
English proficiency
English proficiency improved over time with mean scores 9.82 (SD = 2.80) at Wave 1 to 11.28 (SD = 2.45) at Wave 3 and 12.36 (SD = 2.52) in Wave 5. Table 2 suggests that age at arrival was significantly associated with English proficiency. For each extra year of age at arrival, the English proficiency score reduced by 0.25 points. Compared to age 14, refugees who arrived at 21 years old scored 1.75 points lower in English proficiency (more than half of a standard deviation).
Length of stay was positively associated with English proficiency (p < .05). The improvement was close to a standard deviation between Wave 1 to Wave 3 and Wave 3 to Wave 5. These results suggest that while there was a substantial improvement in English proficiency over time, the age at arrival gap in English proficiency did not change significantly over time. There were several demographic characteristics associated with English proficiency. People who held an Onshore Protection visas prior to resettlement had significantly better English. Refugees born in Iran (p < .1), Iraq, Myanmar and Afghanistan had a poorer English proficiency compared to refugees born in ‘other’ countries. For Iranian and Myanmar refugees there were substantial differences with English proficiency disadvantage more than half a standard deviation lower than refugees born in other countries.
Cultural understanding
The proportion of refugees who found it hard to understand the Australian culture decreased over time, from 45% at Wave 1 to 28% at Wave 3 and to 20% at Wave 5. Table 2 shows that the relationship between age at arrival and cultural understanding was not statistically significant, whereas wave and age by wave were significantly related to cultural understanding. This suggests that there was an overall improvement in cultural understanding over the time but refugees who were older on arrival increasingly reported difficulties in understanding Australian culture. Moreover, females had more difficulties in understanding Australian culture than males.
Belonging and interpersonal trust
There was no statistically significant association between age at arrival, wave and age by wave on belonging and interpersonal trust. Refugees who had Onshore Protection visas reported a greater sense of belonging as did refugees born in Nepal compared to the ‘other countries’ category. The only demographic characteristic associated with interpersonal trust was sex, with females reporting lower trust than males.
Foundations
The proportion of people knowing their rights relatively well were 30%, 56% and 75% at Waves 1, 3 and 5 respectively. Age at arrival, wave and age by wave were significantly related to the knowledge of their own rights. For each extra year of age at arrival, the odds ratio of knowing their rights better was 1.25 times higher. The odds ratios for wave and wave by age suggest a complex relationship and logistic regressions for each wave were used as a supplement to the RE logistic regression models. Based on these models we estimated the relationship between age at arrival and the knowledge of rights at Waves 1, 3 and 5. The probability of knowing their rights well was about 20 percentage points higher among refugees arrived at age 21 compared to refugees arrived at age 14 at Wave 1; however, by Wave 3 the differences by age at arrival were not significant (Figure 2). Figure 2 shows that after five years (Wave 5) those who arrived in Australia at a younger age had greater knowledge of rights than older refugees.
Probability of knowing rights well by age at arrival, Waves 1, 3 and 5
Citation: Longitudinal and Life Course Studies 2023; 10.1332/175795923X16839016657163
Probability of knowing rights well by age at arrival, Waves 1, 3 and 5
Citation: Longitudinal and Life Course Studies 2023; 10.1332/175795923X16839016657163
Probability of knowing rights well by age at arrival, Waves 1, 3 and 5
Citation: Longitudinal and Life Course Studies 2023; 10.1332/175795923X16839016657163
In terms of demographic characteristics, sex and refugees who held Onshore Protection visas prior to resettlement have significant correlation with the knowledge of their rights. The odds of females knowing their rights was 31% lower than males. The odds for refugees who held an Onshore Protection visa prior to arrival to know their rights well was 3.5 times higher than refugee who had the 200 Refugee visa.
Discussion and conclusion
Overall, the findings of this study have indicated that, among refugees who arrived at 14–21 years old in Australia, age at arrival was significantly associated with employment, mental health, social outcomes, knowledge of rights and English language proficiency. Refugees who arrived older tend to have a higher probability of having a paid job but have poorer mental health and English proficiency. In terms of the knowledge of rights, older entrants were more likely to know their rights well compared to younger entrants at arrival, but younger entrants’ knowledge of their rights overtook older entrants four to five years after their arrival. Length of stay in Australia was also significantly related to multiple integration outcomes including clinical levels of PTSD, physical health, employment, English proficiency, cultural understanding, interpersonal trust and knowledge of rights.
As far as we aware, this is the first study of refugee youth that has examined the relationship between age at arrival and all the four core domains of integration identified by Ager and Strang (2008). The findings of this study have shown how the age at arrival relates to the different dimensions of integration outcomes to a greater or lesser extent. Age at arrival was less related to the facilitators compared to the other three core domains of integration with only English proficiency significantly related to age at arrival among refugee youth. As for the length of stay, it was related to all the markers and means measures as well as the knowledge of rights but was not significantly associated with the social outcomes in this study and some of the facilitator measures. In our study some measures for Social Connection and Foundation were quite brief and future research with more comprehensive measures may detect further associations with length of stay.
Markers and means
Health
Prior research suggested that older entrants had higher levels of somatic distress (Brown et al, 2010), PTSD (Mghir et al, 1995) and acculturative stress (Kuo and Roysircar, 2004). In this study older entrants had higher levels of psychological distress, but not necessarily at clinical levels. Moreover, the findings also suggest that the length of stay and age by wave were not significantly associated with both psychological distress and likelihood of having serious mental illness. This is in contrast to the findings from the empirical research on immigrant mental health, where the longer the stay, the lower level of the acculturative stress (Kuo and Roysircar, 2004). While mental health problems were not at clinical levels this finding suggests that unlike other immigrant youth, refugee youth experience ongoing mental health challenges five years post-arrival.
In terms of PTSD, this study’s findings is consistent with Mghir et al (1995) that older entrants were associated with higher risk of PTSD with a larger sample and more diverse sample population. The higher rates of clinical levels of PTSD for older refugee youth could be attributed to greater exposure to trauma prior to arrival; however, the significant increases in Wave 3 suggest older refugees might face more difficulties in settling in Australia compared to younger refugees, such as unemployment and change in social status after resolving two years later (Colic-Peisker, 2003).
Length of stay was significantly related to physical health but not age at arrival. This may be due to access to higher quality healthcare services and medications, the recovering from the physical health problems developed before and after they fled from their country of origin, and better nutrition in Australia. Improvements in physical health were faster among younger refugees compared to their older counterparts. These findings are in contrast to research on the healthy immigrant effect, at least for younger refugees who report significant improvements in health occur after living in the host country for several years (Fuller-Thomson et al, 2011).
Employment
Employment is a key and direct contributor to the well-being of refugees. Previous research has indicated that younger age at arrival has been associated with a higher future earnings, greater likelihood of a stable job and better occupational outcomes (Gonzalez, 2003; Hermansen, 2017). However, in our study, the probability of having a paid job was positively associated with the age at arrival. In the short term, younger refugees will be studying at school or undertaking English courses and training, particularly given that it is not legal to get a job until 15 years of age in Australia. Extra years of training and education in Australia before entering the workforce could result in a better employment outcome and higher earnings in the longer term, as empirical research has shown that the quality of education and years studied in the host country can influence the eventual educational outcomes and hence the employment outcomes (Ager and Strang, 2008; Hermansen, 2017).
Social connections
This study confirms findings in other studies with the younger the age at arrival, the less likely refugee youth were to be socially segregated (Myers et al, 2009; Beck et al, 2012). Younger refugees were at school and had a greater chance to make friends and had better language proficiency (Åslund et al, 2015).
Facilitators
English proficiency
Consistent with many empirical studies, the findings in this study confirm that the older the age at arrival, the lower the English proficiency (Myers et al, 2009; Bleakley and Chin, 2010; Beck et al, 2012; Basu, 2018). This association was evident among refugees who arrived under 18 years of age, and also among older refugee youth. However, this finding contrasts with Correa-Velez et al (2017)’s findings of refugee youth resettled in Melbourne, where age at arrival and language proficiency was not significant. Findings may vary because of differences in the measure of English proficiency or differences in the sample composition. Correa-Velez et al (2017) measured only spoken English, and their sample consisted of 47 refugees who were 10–19 years old at Wave 1, much smaller than in the present study.
We found no evidence of sensitive periods of increased language proficiency. This is consistent with findings that the gap of language acquisition does not change significantly among immigrants who arrived after the sensitive period of language acquisition (after nine years) (Myers et al, 2009; Bleakley and Chin, 2010; Basu, 2018).
Cultural understanding, sense of belonging and interpersonal trust
The findings in this study suggest that the age at arrival and length of stay were not significantly associated with sense of belonging in Australia, possibly due to high levels of belonging. The findings also suggest that age at arrival and the length of stay were not significantly associated with interpersonal trust and the cultural understanding of refugees.
Foundations
The results from this study suggest that in general, the older refugee youth arrive, the better their knowledge of their rights. This could be the result of greater exposure to the migration process, application for refugee status and participation in the decision to migrate. However, the relationship between age at arrival and the knowledge of rights reversed over time. Young refugees who arrived with relatively poor understanding of their rights gained in their knowledge of rights more quickly than older refugee youth. This is likely to be due to greater improvements in English proficiency and more support and access to information at school. It is notable older entrants’ knowledge of their own rights changed little from first arrival. As rights are the foundation of integration, knowing the responsibilities and rights of citizens could help refugees better understand the Australian identity and lead to a greater sense of belonging and sense of security (Ager and Strang, 2008). Policy makers should consider providing more assistance and information of rights, such as information sessions, to refugee youth who arrive at an older age.
Sex differences and integration
Female youth had worse integration outcomes, with females having poorer mental and physical health, social and employment outcomes, cultural understanding, interpersonal trust and the knowledge of rights than their male counterparts. In this study there was no integration outcome where females were better than males and provision of greater support to female refugee youth is a policy priority.
Age at arrival and integration of youth and emerging adult refugees
This study also contributes to our understanding of emerging adult refugees’ integration into Australia. During this period, emerging adults are more likely to be struggling with their identities, feeling displaced and distressed because of the uncertainty of future (Arnett, 2000; 2004; Arnett et al, 2014). Disadvantages in cultural understanding, language and social integration could further hinder the identity exploration of emerging adult refugees in host countries. However, having a paid job might help emerging adult refugees to develop their identity and sense of security. As for the mental health outcomes, emerging adult refugees have poorer mental health than younger refugee youth which make it more difficult to integrate and exacerbate the stressors associated with the period of emerging adulthood.
Limitations
There are several limitations to this study that need to be noted. First, this study could not capture the relationships between the age at arrival and the long-term integration outcomes, such as school completion and marital outcomes, with a five-year longitudinal study. As we have already mentioned, integration is a complex and enduring process: the observation period in this research might not be sufficient for us to study the long-term integration outcomes. However, using five years of BNLA data, this study indicates how the age at arrival could influence short-term integration outcomes, which are important, as early settlement experiences can potentially affect the long-term trajectories of refugees. Second, parental effects and traumatic experiences were not examined in this study (Lester and Flake, 2013; Bryant et al, 2018). Third, there was no immigrant reference group to test whether the relationships were refugee-specific. Finally, because of the limitation of data and the selection of sample, integration outcomes in some dimensions could not be measured. For instance, the BNLA survey does not enable us to robustly measure the relationship between age at arrival and education and training outcomes. Therefore, education was not measured in this study. However, despite the limitations, the findings in this longitudinal study have contributed to the understanding of the effect of age at arrival on refugee integration by studying the relationship between age at arrival and length of stay on refugee multidimensional integration outcomes using the five years of longitudinal data from refugee youth.
Empirical research has indicated that age at arrival has significant effects on different integration outcomes. This study offers new evidence on the relationship between age at arrival and multiple integration outcomes. Building on the theories of refugee integration and the age at arrival, we examined the relationships between the age at arrival and the integration outcomes based on the framework of integration developed by Ager and Strang (2008). Overall, the findings reveal that younger refugees had better integration outcomes. This study reaffirms that integration is a complex and multidimensional process that cannot be fully captured by only examining specific dimension of integration outcomes. There are numerous avenues for further research, testing whether parental factors mediate age at arrival, examining intercorrelation of integration outcomes over time, and examining refugee youth cohorts beyond five years from arrival. A better understanding of the multidimensionality of the integration process and the relationship between age at arrival and integration can help the policy makers and local communities to provide better and more appropriate supports to refugees in different life stages. Better integration not only can benefit refugees, but also the community as a whole, as successful integration can enhance social harmony and socio-cultural diversity in the host community.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Data access
Detail for data access can be found in the Australian Data Archive website: https://dataverse.ada.edu.au/dataverse.xhtml?alias=bnla
Conflict of interest
The authors declare that there is no conflict of interest.
References
Ager, A. and Strang, A. (2008) Understanding integration: a conceptual framework, Journal of Refugee Studies, 21(2): 166–91. doi: 10.1093/jrs/fen016
Arnett, J.J. (2000) Emerging adulthood: a theory of development from the late teens through the twenties, American Psychologist, 55(5): 469–80. doi: 10.1037/0003-066X.55.5.469
Arnett, J.J. (2004) Emerging Adulthood: The Winding Road from the Late Teens Through the Twenties, Oxford: Oxford University Press.
Arnett, J.J., Žukauskienė, R. and Sugimura, K. (2014) The new life stage of emerging adulthood at ages 18–29 years: implications for mental health, The Lancet. Psychiatry, 1(7): 569–76. doi: 10.1016/S2215-0366(14)00080-7
Åslund, O., Böhlmark, A. and Skans, O.N. (2015) Childhood and family experiences and the social integration of young migrants, Labour Economics, 35: 135–44.
Australian Bureau of Statistics (2012) Information Paper: Use of the Kessler Psychological Distress Scale in Abs Health Surveys, Australia, 2007–08, https://www.abs.gov.au/ausstats/abs@.nsf/Lookup/4817.0.55.001Chapt er92007- 08#:~:text=K6%20SCORING%20%26%20CATEGORISATION,maxi mum%20possible%20score%20of%2024.
Basu, S. (2018) Age-of-arrival effects on the education of immigrant children: a sibling study, Journal of Family and Economic Issues, 39(3): 474–93. doi: 10.1007/s10834-018-9569-4
Beck, A., Corak, M. and Tienda, M. (2012) Age at immigration and the adult attainments of child migrants to the United States, ANNALS of the American Academy of Political and Social Science, 643(1): 134–59. doi: 10.1177/0002716212442665
Bleakley, H. and Chin, A. (2010) Age at arrival, English proficiency, and social assimilation among US immigrants, American Economic Journal: Applied Economics, 2(1): 165–92. doi: 10.1257/app.2.1.165
BNLA (Building a New Life in Australia) (2019) BNLA Data User Guide, Release 5.0, Melbourne, VIC: Australian Institute of Family Studies.
Böhlmark, A. (2008) Age at immigration and school performance: a siblings analysis using Swedish register data, Labour Economics, 15(6): 1366–87.
Bornstein, M.H. (2012) Cultural approaches to parenting, Parenting, Science and Practice, 12(2/3): 212–21. doi: 10.1080/15295192.2012.683359
Bosker, R. and Snijders, T.A. (2011) Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling, London: SAGE.
Brown, C., Schale, C.L. and Nilsson, J.E. (2010) Vietnamese immigrant and refugee women’s mental health: an examination of age of arrival, length of stay, income, and English language proficiency, Journal of Multicultural Counseling and Development, 38(2): 66–76. doi: 10.1002/j.2161-1912.2010.tb00115.x
Bryant, R.A., Edwards, B., Creamer, M., O’Donnell, M., Forbes, D., Felmingham, K.L. et al. (2018) The effect of posttraumatic stress disorder on refugees’ parenting and their children’s mental health: a cohort study, The Lancet Public Health, 3(5): e249–e258, doi: 10.1016/S2468-2667(18)30051-3.
Cohen Goldner, S. and Epstein, G.S. (2014) Age at immigration and high school dropouts, IZA Journal of Migration, 3: art 19, doi: 10.1186/s40176-014-0019-2.
Colic-Peisker, V. (2003) European Refugees in (White) Australia: Identity, Community and Labour Market Integration, National Europe Centre Paper No. 111, Canberra: ANU Centre for European Studies, https://openresearch-repository.anu.edu.au/bitstream/1885/41500/3/vpeiskerNECpaper.pdf.
Correa-Velez, I., Gifford, S.M., McMichael, C. and Sampson, R. (2016) Predictors of secondary school completion among refugee youth 8 to 9 years after resettlement in Melbourne, Australia, Journal of International Migration and Integration, 18(3): 791–805. doi: 10.1007/s12134-016-0503-z
Cortes, K.E. (2006) The effects of age at arrival and enclave schools on the academic performance of immigrant children, Economics of Education Review, 25(2): 121–32. doi: 10.1016/j.econedurev.2004.12.001
Edwards, B., Smart, D., De Maio, J., Silbert, M. and Jenkinson, R. (2018) Cohort profile: Building a New Life in Australia (BNLA): the longitudinal study of humanitarian migrants, International Journal of Epidemiology, 47(1): 20–20h. doi: 10.1093/ije/dyx218
Fadnes, L.T. and Diaz, E. (2017) Primary healthcare usage and use of medications among immigrant children according to age of arrival to Norway: a population-based study, BMJ Open, 7(2): e014641, doi: 10.1136/bmjopen-2016-014641.
Fuller-Thomson, E., Noack, A.M. and George, U. (2011) Health decline among recent immigrants to Canada: findings from a nationally-representative longitudinal survey, Canadian Journal of Public Health, 102(4): 273–80. doi: 10.1007/BF03404048
Gonzales, A. (2003) The education and wages of immigrant children: the impact of age at arrival, Economics of Education Review, 22(2): 203–12. doi: 10.1016/S0272-7757(02)00004-3
Hansen, M., Andersen, T.E., Armour, C., Elklit, A., Palic, S. and Mackrill, T. (2010) PTSD-8: a short PTSD inventory, Clinical Practice and Epidemiology in Mental Health, 6(1): 101–8, doi: 10.2174/1745017901006010101.
Hermansen, A.S. (2017) Age at arrival and life chances among childhood immigrants, Demography, 54(1): 201–29. doi: 10.1007/s13524-016-0535-1
Home Office (2019) Home office indicators of integration framework 2019, home office research report 109, https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1074688/home-office-indicators-of-integration-framework-2019-horr109.pdf.
Kuo, B.C.H. and Roysircar, G. (2004) Predictors of acculturation for Chinese adolescents in Canada: age of arrival, length of stay, social class, and English reading ability, Journal of Multicultural Counseling and Development, 32(3): 143–54. doi: 10.1002/j.2161-1912.2004.tb00367.x
Lemmermann, D. and Riphahn, R.T. (2018) The causal effect of age at migration on youth educational attainment, Economics of Education Review, 63: 78–99. doi: 10.1016/j.econedurev.2017.11.001
Lester, P. and Flake, E. (2013) How wartime military service affects children and families, The Future of Children, 23(2): 121–41. doi: 10.1353/foc.2013.0015
Lichtenstein, G. and Puma, J.E. (2019) The refugee integration survey and evaluation (RISE): results from a four-year longitudinal study, Journal of Refugee Studies, 32(3): 397–416. doi: 10.1093/jrs/fey034
Mghir, R., Freed, W., Raskin, A. and Katon, W. (1995) Depression and posttraumatic stress disorder among a community sample of adolescent and young adult Afghan refugees, Journal of Nervous and Mental Disease, 183(1): 24–30. doi: 10.1097/00005053-199501000-00005
Myers, D., Gao, X. and Emeka, A. (2009) The gradient of immigrant age-at-arrival effects on socioeconomic outcomes in the U.S., The International Migration Review, 43(1): 205–29. doi: 10.1111/j.1747-7379.2008.01153.x
Pak, S. (2010) The growth status of North Korean refugee children and adolescents from 6 to 19 years of age, Economics & Human Biology, 8(3): 385–95.
Pritchard, P., Maehler, D.B., Pötzschke, S. and Ramos, H. (2019) Integrating refugee children and youth: a scoping review of English and German literature, Journal of Refugee Studies, 32(S1): i194–i208. doi: 10.1093/jrs/fez024
UNHCR (United Nations High Commissioner for Refugees) Australia (2020) Global trends: forced displacement in 2019, https://www.unhcr.org/en-au/statistics/unhcrstats/5ee200e37/unhcr-global-trends-2019.html.
van Ours, J.C. and Veenman, J. (2006) Age at immigration and educational attainment of young immigrants, Economics Letters, 90(3): 310–16. doi: 10.1016/j.econlet.2005.08.013
