Understanding access to higher education amongst humanitarian migrants: an analysis of Australian longitudinal survey data

Abstract

Humanitarian migrants are amongst the most marginalised population groups in countries within the Global North, including Australia. An important channel for these migrants to successfully settle into the host society and improve their socio-economic outcomes is participation in the local education system, particularly in higher-education options. However, we know surprisingly little about the socio-demographic factors that structure inequalities in humanitarian migrants’ access to (higher) education, with evidence from robust quantitative studies being particularly scarce. The present study fills this important gap in knowledge by analysing Australian longitudinal survey data (Building a New Life in Australia; n = 2109 migrants and 8668 person-year observations) by means of random-effect panel regression models. Key results indicated that higher English-language proficiency and pre-arrival education levels are core factors fostering greater engagement with the Australian higher-education system amongst humanitarian migrants. Humanitarian-migrant women in our sample exhibited a greater adjusted likelihood of being a student than humanitarian-migrant men. Altogether, our findings confirmed inequalities in accessing the Australian higher-education system amongst humanitarian migrants, and that policy attention is required to redress this situation. However, they also stress that a ‘one size fits all’ policy strategy may be neither sufficient nor appropriate to boost their education prospects.

Full text

Vol.:(0123456789)
Higher Education
https://doi.org/10.1007/s10734-021-00772-x
1 3
Understanding access tohigher education
amongsthumanitarian migrants: ananalysis ofAustralian
longitudinal survey data
FranciscoPerales1· NingXiang2· LisaHartley3· MatthiasKubler2·
WojtekTomaszewski2
Accepted: 30 September 2021
© The Author(s), under exclusive licence to Springer Nature B.V. 2021
Abstract
Humanitarian migrants are amongst the most marginalised population groups in countries
within the Global North, including Australia. An important channel for these migrants to
successfully settle into the host society and improve their socio-economic outcomes is par-
ticipation in the local education system, particularly in higher-education options. However,
we know surprisingly little about the socio-demographic factors that structure inequalities
in humanitarian migrants’ access to (higher) education, with evidence from robust quanti-
tative studies being particularly scarce. The present study fills this important gap in knowl-
edge by analysing Australian longitudinal survey data (Building a New Life in Australia;
n = 2109 migrants and 8668 person-year observations) by means of random-effect panel
regression models. Key results indicated that higher English-language proficiency and pre-
arrival education levels are core factors fostering greater engagement with the Australian
higher-education system amongst humanitarian migrants. Humanitarian-migrant women
in our sample exhibited a greater adjusted likelihood of being a student than humanitar-
ian-migrant men. Altogether, our findings confirmed inequalities in accessing the Austral-
ian higher-education system amongst humanitarian migrants, and that policy attention is
required to redress this situation. However, they also stress that a ‘one size fits all’ policy
strategy may be neither sufficient nor appropriate to boost their education prospects.
Keywords Australia· Education· Equity· Humanitarian migrants· Refugees· University
* Francisco Perales
f.perales@uq.edu.au
1 School ofSocial Science, The University ofQueensland, Brisbane, QLD, Australia
2 Institute forSocial Science Research, The University ofQueensland, Brisbane, QLD, Australia
3 Centre forHuman Rights Education, Curtin University, Perth, WA, Australia

Higher Education
1 3
Introduction
According to the 1951 United Nations Convention on the Status of Refugees and its 1967
Protocol, a refugee is any person who owing to a well-founded fear of being persecuted for
reasons of race, religion, nationality, membership of a particular social group or political
opinion, is outside the country of his/her nationality and is unable, or owing to such fear,
unwilling to avail themselves to the protection of that country. Humanitarian migrants are
refugees who have applied for asylum and have been granted protection, whereas asylum
seekers are people who have sought protection but whose claims have not been finalized
(OECD, 2016). The current study focuses on humanitarian migrants to Australia, specifi-
cally on those who were granted permanent protection visas through Australia’s humani-
tarian programmes during May 2013 and December 2013. These migrants are treated as
permanent residents of Australia and enjoy the same services and supports as Australian
citizens and other permanent residents. This includes the rights to work, receive income-
support payments and participate in education. When entering higher education (HE),
humanitarian migrants in Australia also enjoy the same levels of fee support as citizens and
other permanent residents.
Despite the support offered by government institutions and third-sector organiza-
tions, humanitarian migrants remain amongst the most marginalised population groups in
Australia (e.g. Hugo, 2011), similar to other countries across the Global North (OECD,
2016, 2019). For example, the employment rates of recently arrived working-age humani-
tarian migrants in Australia, 33.3% for men and 7.5% for women (DSS, 2017), are strik-
ingly lower than those of the Australian-born population, 81.1% for men and 69.6% for
women (Wilkins & Lass, 2018). This is consistent with global evidence showing that many
humanitarian migrants around the world are unemployed, under-employed or work in jobs
that do not match their skills (OECD, 2016). An important channel to enable humanitar-
ian migrants to settle into the host society and improve their socio-economic outcomes is
participation in the local education system. Attainment of home-country educational cre-
dentials—particularly higher-education credentials—not only creates employment path-
ways for humanitarian migrants; it also improves their cultural-competency skills, widens
their social networks and enhances their subjective wellbeing (Cerna, 2019; Streitwieser
etal., 2018; UNHCR, 2019). Access to HE may be particularly beneficial to humanitar-
ian migrants who are younger or early into their careers, hold tertiary education aspira-
tions or lack sufficient credentials to find suitable work in Australia. However, humani-
tarian migrants face multiple and unique challenges to accessing and participating in the
host-country HE system, both in Australia (Baker etal., 2019a, 2019b; Correa-Velez etal.,
2015; Harris & Marlowe, 2011; Hartley etal., 2018) and internationally (e.g., Bajwa etal.,
2017; Cerna, 2019). For these reasons, scholars and policymakers have called for further
support for this group to engage in HE (Bajwa etal., 2017; Baker etal., 2019a, b; Cerna,
2019; Delaporte & Piracha, 2018; Hartley etal., 2018).
Policy responses aimed at lifting the educational engagement of humanitarian migrants
should be based on a robust body of empirical evidence. While qualitative research has pro-
vided valuable insights into the experiences of humanitarian migrants in the host-society’s
education system, quantitative studies on this topic remain scarce (Cerna, 2019; Naidoo
etal., 2018; Ramsay & Baker, 2019; Streitwieser etal., 2018; Terry etal., 2016). General-
izable, quantitative evidence on the associations between humanitarian migrants’ circum-
stances and access to HE would further help guide and develop programs and interventions
aimed at improving their educational engagement. Because of the greater economic and

Higher Education
1 3
social rewards that stem from participation in HE (see, e.g. ABS, 2017; Desjardins & Lee,
2016), it is particularly significant for new studies to consider inequalities to accessing HE.
This study fills this gap in knowledge. Specifically, it sets out to identify the factors that
correlate with disparities in access to education and HE amongst humanitarian migrants.
In doing so, the findings can help design programs and policies that equalize access and
address the distinct needs and challenges that different groups of humanitarian migrants
may face. To this aim, we present the results of analyses of longitudinal survey data exam-
ining the socio-demographic factors that structure access to education and HE amongst
a recent cohort of humanitarian migrants to Australia. The analyses consider also differ-
ences between men and women, thereby recognising the potential gendering of these pro-
cesses (Harris et al., 2015; Hatoss & Huijser, 2010; Seck, 2015). Australia is an interest-
ing case study to examine access to HE amongst humanitarian migrants; it is a traditional
destination country for humanitarian migrants from across the globe and it offers many of
these immigrants (including all of those captured in our study) equal rights to education
as the local-born. Our key results indicate that higher English-language proficiency and
pre-arrival education levels are core factors fostering greater engagement with the Austral-
ian higher-education system amongst humanitarian migrants. Additionally, humanitarian-
migrant women exhibited a greater adjusted likelihood of being a student than humanitar-
ian-migrant men.
Literature review
Conceptual framework: humanitarian‑migrant status andinequalities inaccess
toHE
Despite recent efforts to increase humanitarian migrants’ education access and participa-
tion in countries within the Global North (e.g. Mangan & Winter, 2017; Manhica etal.,
2019; Morrice, 2009; Ramsay & Baker, 2019), the educational status of this population
remains a matter of concern (Streitwieser etal., 2018; UNHCR, 2019). Much of the lit-
erature examining inequalities in access to education amongst humanitarian migrants has
focused on access to and success in primary and secondary education (Streitwieser etal.,
2018). However, a growing body of studies is beginning to explore humanitarian migrants’
access to and participation in HE (Ramsay & Baker, 2019).
In conceptualising humanitarian-migrant access to HE, it is important to consider two
types of underlying inequalities: between- and within-group inequalities. First, there are
significant access inequalities between humanitarian migrants and other population groups
(including other immigrants and native-born individuals). While there is little quantitative
evidence on the HE access rates of humanitarian migrants in Australia (cf. Terry et al.,
2016), we know from qualitative studies that this group faces significant challenges in pur-
suing HE (see, e.g. Hartley et al., 2018; Joyce et al., 2010; Stevenson & Baker, 2018).
Further, statistical evidence indicates that humanitarian-migrants’ average educational
attainment is significantly lower than for other groups. For example, our own calculations
revealed that, in the BNLA data, just 11.5% of humanitarian migrants aged 20–65years
had completed a degree at the point of arrival. Further, less than 10% of these degrees
had been partly or fully recognised in Australia in the first 5 years post-settlement. In
comparison, our calculations using data from the 2019 Household, Income and Labour
Dynamics in Australia Survey (Summerfield etal., 2020) revealed that 32% of the overall

Higher Education
1 3
Australian population and 44% of immigrants in Australia aged 20–65years held degree-
level qualifications.
Second, in conceptualising access to HE, we must acknowledge existing inequalities
within humanitarian migrants. In thinking about such within-group inequalities—which
constitute the focus of the present study—some scholars have found it useful to refer to
the notion of ‘barriers to education’ (see, e.g. Bajwa etal., 2017; Hatoss & Huijser, 2010;
Watkins etal., 2012). Within this context, the term ‘barriers’ denotes systemic factors out
of the control of the individual that make it harder for some subgroups of humanitarian
migrants to access HE. Barriers to HE are multidimensional (i.e. they tap into multiple life
domains) and unequally experienced across subgroups of humanitarian migrants (e.g. by
gender or socio-economic background) (e.g. Hartley etal., 2018). This unequal distribution
means that some socio-demographic factors may act as empirical correlates of HE access
amongst humanitarian migrants, as we elaborate below.
Factors structuring access toHE amongsthumanitarian migrants toAustralia
Previous studies point to factors that may structure access to HE amongst humanitarian
migrants (Bajwa etal., 2017; Hatoss & Huijser, 2010; Streitwieser etal., 2018; UNHCR,
2019). These include health status, information barriers and material/financial conditions.
Concerning health status, research indicates that humanitarian migrants often experience
health sequelae stemming from traumatic experiences of persecution, violence or reten-
tion—including mental-health issues such as post-traumatic stress disorder (see, e.g. Slewa-
Younan etal., 2015). In addition, some humanitarian-migrant youth exhibit delays in their
cognitive and socio-emotional developmental trajectories due to spending long portions
of their childhood in camps or in transit (Cerna, 2019). This suggests that humanitarian
migrants’ health status should be a significant correlate of their participation in HE. Like-
wise, we may observe greater levels of HE access amongst humanitarian migrants with less
complex migration pathways, such as ‘onshore’ compared to ‘offshore’ applicants.
Information barriers refer to difficulties accessing professional support and navigating
educational pathways (Bajwa etal., 2017; Morrice, 2009). For example, Morrice (2009)
documents that individuals from humanitarian-migrant backgrounds found it hard to access
accurate information that helps them navigate the UK educational system. Similarly, Bajwa
etal. (2017) found that humanitarian-migrant students in Canada struggled to find useful
information on how to have their credentials assessed, finance their HE studies or utilise
HE online resources. Based on this, we may expect that those humanitarian migrants who
are better equipped with skills to navigate HE admission and funding processes—and HE
life itself—may be more likely to access HE. This may apply, for instance, to humanitarian
migrants who had attained greater education levels prior to arriving in Australia, or those
who have greater levels of English-language proficiency. Similarly, we may expect humani-
tarian migrants to be progressively more likely to enrol in HE the longer they are in the
host country, as they develop their social networks and become more familiar with Austral-
ian educational, financial and other institutions.
Material and financial resources are also relevant, as they shape humanitarian
migrants’ ability to pay for education-related tuition, materials, transportation and
other fees (UNHCR, 2019). Financial challenges may also emerge when humanitarian
migrants need to prioritise family needs over HE—for instance, when they must engage
in paid employment to support their nuclear or extended family instead of undertak-
ing HE studies (UNHCR, 2019). Those humanitarian migrants who are not in paid

Higher Education
1 3
employment may thus be more likely to engage in HE, due to a greater time availability.
Spatial inequalities in access to HE may also be important in the Australian context,
with humanitarian migrants residing in major cities being better able to access HE than
those residing in regional areas because of a greater availability of HE institutions and
pathways.
In addition to these key factors, other aspects may also structure humanitarian
migrants’ access to HE in the host country. These include age—with younger humani-
tarian migrants expecting greater long-term returns to education, and hence more likely
to select into HE (see, e.g. OECD, 2019); country background—where different aggre-
gate-level country-of-origin influences may foster or inhibit HE access (see, e.g. Joyce
etal., 2010); and work experience—with humanitarian migrants with work experience
being better able to secure paid work without the need to upskill, and hence less likely
to participate in HE (see, e.g. Baker etal., 2019a, b).
Gender inequalities inHE access amongsthumanitarian migrants
Gender is a powerful force structuring access to opportunities within society, as well as
the attainment of valued socio-economic outcomes. A wealth of literature has documented
gender-based inequalities amongst immigrants—not just humanitarian migrants—upon
arrival in a host society. These studies often find that immigrant women tend to enjoy fewer
opportunities for social and economic participation than immigrant men (see, e.g. ABS,
2020; Krieger, 2020; OECD-European Union, 2018). These patterns have been linked to
traditional gender ideologies underpinning migration decision-making processes, as well
as post-settlement decisions around spousal economic specialization in paid or unpaid
work (Krieger, 2020; OECD-European Union, 2018). Whether and how these gendered
processes manifest amongst humanitarian migrants is less clear.
Some authors have documented that humanitarian-migrant women are more dis-
advantaged than humanitarian-migrant men in their ability to pursue HE in the host-
country society (e.g. Harris et al., 2015; Hatoss & Huijser, 2010; Seck, 2015). Draw-
ing on interviews with 14 Sudanese humanitarian migrants who had recently arrived in
Australia, Hatoss and Huijser (2010) demonstrated that cultural-gender factors played a
key role in restricting opportunities for women to pursue HE by locating women’s roles
within the family home. Similar processes have been documented for Karen humanitar-
ian-migrant women in Australia (Watkins etal., 2012). Consistent with this, Terry etal.
(2016) reported that the percentage of humanitarian-migrant women students in Aus-
tralia was smaller than that of men, although it had increased from 30% in 2009 to 40%
in 2014. However, this percentage varied markedly by country of origin—suggesting
that cultural norms may play a role in determining humanitarian-women’s access to HE.
Besides this gender ‘main effect’, there may also be gender-asymmetrical associations
between some socio-demographic factors and humanitarian migrants’ likelihood of access-
ing HE. Specifically, given findings for other immigrants (e.g. ABS, 2020; OECD-Euro-
pean Union, 2018) and evidence of traditional gender roles amongst some subgroups of
humanitarian migrants (e.g. Harris etal., 2015; Hatoss & Huijser, 2010), we expect certain
factors to inhibit humanitarian-migrant women’s access to HE. These factors include being
in a marriage—or marriage-like relationship—and having dependent children. A novel
contribution of this study is to empirically examine how gender affects HE access amongst
humanitarian migrants—both as a direct effect and a moderator—using unique survey data.

Higher Education
1 3
The Australian evidence base
A mature body of Australian research has deployed qualitative methods to understand
inequalities in humanitarian migrants’ access to and success in HE (e.g. Baker etal.,
2019a, b; Earnest et al., 2010; Harris & Marlowe, 2011; Harris et al., 2015; Hartley
etal., 2018; Hatoss & Huijser, 2010; Joyce etal., 2010; Naidoo, 2019). Consistent with
the international literature, these studies report that humanitarian migrants face chal-
lenges in accessing and participating in the HE system, including strenuous family and/
or financial responsibilities while undertaking their studies (e.g. supporting family in
Australia or back home) (Joyce et al., 2010), challenges adapting to new educational
contexts and community expectations (Harris & Marlowe, 2011) and inadequate HE
support systems that fail to recognise their unique needs (Earnest etal., 2010). How-
ever, these studies tend to focus on the experiences of humanitarian migrants that have
already entered the Australia HE system, rather than on access inequalities at the point
of entry. As an exception, Stevenson and Baker (2018) discussed factors structuring
humanitarian migrants’ interactions with the HE system along three stages of the stu-
dent life cycle: HE access, HE participation and transitions out of HE. Some factors,
such as (un)interrupted educational experiences and past and ongoing trauma, exerted
negative impacts throughout all three stages. Other factors were particularly relevant
for HE access, including school performance, English-language proficiency, access to
information on HE pathways and receipt of targeted support (Stevenson & Baker, 2018).
In contrast to this well-developed qualitative body of evidence, only one quantitative
study has examined access to HE amongst humanitarian migrants in Australia (Terry
etal., 2016). Using 2009–2014 administrative data, Terry etal. (2016) reported that the
number of students from humanitarian-migrant backgrounds enrolling in Australian HE,
while small, had increased from 1687 (0.21% of all domestic students) in 2009 to 3506
in 2014 (0.34% of all domestic students). This study, however, did not engage with other
correlates of humanitarian-migrant access to, or participation in, HE.
The current study: aims andcontributions
This study makes a significant contribution to the literature regarding the factors that
structure access to education and HE amongst humanitarian migrants by redressing sev-
eral of its current shortcomings. A first limitation of the available evidence base is that
most studies follow a qualitative methodological approach, both internationally and in
Australia (Cerna, 2019; Ramsay & Baker, 2019). Ramsay and Baker (2019) reviewed
46 journal articles and noted ‘a strong commitment to qualitative inquiry in the field’
(p.79). Specifically, all 32 empirical articles included in the review relied on qualita-
tive methods. While some studies collected data via mixed methods (typically a small-
scale survey in conjunction with individual interviews or focus groups), the quantita-
tive data collected in these studies were not representative and rarely analysed using
rigorous statistical techniques (Ramsay & Baker, 2019). The present study contributes
to the evidence base by undertaking robust quantitative analyses of a large-scale survey,
providing results that can be generalised to a recent cohort of humanitarian migrants
to Australia. Further, the richness of the survey data at hand allows us to examine
the role of a variety of individual- and family-level factors, offer novel insights into

Higher Education
1 3
humanitarian migrants’ educational trajectories and pay attention to the potentially dis-
tinct ways in which humanitarian migrant men and women access education and HE.
Second, most of the available evidence concerning humanitarian migrants’ interactions
with the HE system has focused on humanitarian migrants who had already been admitted
into universities (e.g. Baker etal., 2019a, b; Cerna, 2019; Joyce etal., 2010; Naidoo, 2019).
Hence, their findings are most relevant to identifying factors structuring success in—rather
than access to—HE. The present study will fill this gap in knowledge by considering the
socio-demographic factors of humanitarian migrants that are associated with increasing or
decreasing odds of accessing education, and HE more specifically.
Data andmethods
Dataset andsample
Building a New Life in Australia: The Longitudinal Study of Humanitarian Migrants
(BNLA) is an internationally distinctive, longitudinal study of humanitarian migrants
(Department of Social Services, 2017). The study has interviewed a sample of 2399
humanitarian migrants from 1510 households within Australia on an annual basis between
2013/2014 (Wave 1) and 2017/2018 (Wave 5). The in-scope population for the BNLA
study comprises adult humanitarian migrants settling in Australia with a permanent visa
between May and October 2013, with the sample selected using complex probabilis-
tic methods (AIFS, 2018). The study collects information from two types of humanitar-
ian migrants: (i) ‘offshore migrants’ who received a permanent humanitarian visa over-
seas and arrived in Australia between May 2013 and December 2013 and (ii) ‘onshore
migrants’ who sought asylum after arriving in Australia and were subsequently granted a
permanent humanitarian visa between May 2013 and December 2013 (AIFS, 2018, p. 3).
All survey materials were translated into the respondents’ mother-tongue or preferred lan-
guages and bilingual interviewers and interpreters were engaged as required. The study’s
initial response rate was approximately 55%, with the subsequent wave-on-wave response
rates being approximately 80%. Our analyses of BNLA data are based on the subsample of
responding individuals of working age (18 to 64years),1 encompassing 8668 observations
from 2109 individuals.
Measures
The outcome variables of interest are two binary variables capturing whether at the time of
the interview the respondent was studying a course at an Australian educational institution
(1 = yes, 0 = no). We consider both: (i) taking part in any type of course and (ii) taking part
in a HE course (at ISCED level 6 +). The latter involves differentiating between University
degrees (1 = yes) and trade or technical qualifications, paid traineeships, work experience,
secondary school (grades 7–12) and short courses (0 = no).
Our explanatory variables capture a range of factors known to affect individuals’ deci-
sions concerning their economic and education participation. We selected explanatory
1 While this is a broad age range, our results were generally robust to restricting our sample to individuals
aged younger than 51years or younger than 31years (see Appendix 1 Table5).

Higher Education
1 3
variables that approximate the socio-demographic and economic factors structuring access
to HE amongst humanitarian migrants discussed before; that have been considered in pre-
vious studies of (humanitarian-)migrant HE access; or that could confound the relation-
ships of interest. These include continuous measures of respondent’s age (expressed in
years) and its square (to capture non-linear relationships), self-assessed spoken-English
proficiency, self-assessed general health and number of children in the household; binary
variables capturing whether the respondent is a woman, migrant type (onshore/offshore),
area of residence (regional/non-regional), current employment status and being ever in paid
work prior to arriving in Australia; and sets of dummy variables capturing respondent’s
highest educational qualification prior to arriving in Australia, partnership status, length of
time in Australia; and country of birth. Appendix 1 Table3 presents means and standard
deviations for all analytic variables, pooling observations from survey participants across
time points. A more detailed breakdown of the characteristics of humanitarian migrants
who participate in education and HE can be found in Appendix 1 Table4.
Analytic approach
The aim of our analyses is to examine the patterns and predictors of humanitarian-
migrants’ engagement with the Australian education system. To accomplish this, we lever-
age the longitudinal nature of the BNLA data in two main ways. First, we use these data
to document gender-specific trajectories in access to education over the first 5years post-
settlement. Second, we exploit the panel structure of the BNLA dataset to fit random-effect
logistic regression models predicting humanitarian-migrants’ participation in education
and HE (for details, see Appendix 2). These models account for the nesting of observa-
tions within the same individuals and leverage the panel data to improve the estimation
of the relationships of interest through the incorporation of an individual-specific random
intercept (Wooldridge, 2010). The analyses use random-effect logistic regression models
because the outcomes capturing access to education are all dichotomous variables. For ease
of interpretation, model coefficients are expressed as odds ratios (ORs). We subsequently
examine gender differences in access to education by fitting analogous models in which all
of the explanatory variables were interacted with a ‘woman’ dummy variable.
Empirical evidence
Levels ofandtrends ineducational engagement
The top of Table3 presents sample averages for the educational variables of interest. Pool-
ing all time periods, 12.6% of humanitarian migrants were studying towards a qualifica-
tion in Australia; 1.7% were studying towards a degree, 5% a trade/technical course or a
paid traineeship and a further 5.3% another course or work-experience program.2 That is,
around 14.2% of all students were studying towards a degree. These figures pertain to the
overall sample and do not yield any insights into temporal trends. The results presented
2 The percentages for the different education categories do not add up to the overall percentage because a
small number of respondents reported undertaking multiple courses falling into more than one of the cat-
egories (e.g. 5.3 + 5.0 + 1.7 ≠ 12.6). This also applies to other education variables.

Higher Education
1 3
in Table1, however, are split by survey wave. The pooled figures in Panel A reveal that
the percentage of humanitarian migrants who were studying any course increased between
wave 1 (10.5%) and wave 3 (15.4%), but decreased thereafter (11.2% in wave 5). The per-
centage of humanitarian migrants who studied towards a degree, however, increased lin-
early over time: from 0.5% in wave 1 to 2.6% in wave 5.
The second and third panels present the figures for men (panel B) and women (panel
C) separately. Across survey waves, more humanitarian-migrant women (14.3%) than men
(11.2%) were current students, with some evidence of gender differences in time trends.
For example, the share of women who were current students raised from 10.7% in wave
1 to 14.4% in wave 5, while a decrease was observed for men (from 10.3 to 8.4%). The
increase in the share of women accessing HE (0.8% in wave 1 and 3.6% in wave 5) was
also much more pronounced than the equivalent increases for men (0.4% in wave 1 and
1.7% in wave 5).
Predictors ofengagement ineducation
The results of random-effect logistic regression models examining the socio-demographic
factors associated with being a current student in Australia are shown in columns (1) and
(2) in Table2. The results for being a current student in any educational stream are pre-
sented in column (1). All else being equal, the following factors significantly increased
the odds of being a current student in Australia: being a woman (OR = 1.54, p < 0.01);
higher levels of spoken English-language proficiency (OR = 2.43, p < 0.01); better health
(OR = 1.12, p < 0.01); being single, rather than married/partnered (OR = 2.43, p < 0.01);
and coming from Afghanistan (OR = 1.44, p < 0.05), Iran (OR = 2.48, p < 0.01), Myan-
mar (OR = 5.22, p < 0.01) or another country (OR = 1.76, p < 0.01), rather than from Iraq.
Meanwhile, other factors significantly decreased the odds of being a current student, cet-
eris paribus: having already completed an Australian qualification (OR = 0.18, p < 0.01);
having no educational qualifications (OR = 0.36, p < 0.01), some schooling (OR = 0.46,
p < 0.01) or a trade qualification (OR = 0.63, p < 0.1), compared to a degree, prior to arriv-
ing in Australia; being an onshore rather than an offshore migrant (OR = 0.59, p < 0.01);
Table 1 Sample means for
outcome variables, by survey
wave
BNLA, waves 1–5 (2013–2017)
Mean/%
Wave 1 Wave 2 Wave 3 Wave 4 Wave 5
Panel A: all respondents
Any course 10.5 14.1 15.4 12.4 11.2
Degree course 0.5 1.2 2.2 2.5 2.6
n (observations) 2079 1695 1585 1609 1568
Panel B: men
Any course 10.3 13.4 13.0 10.9 8.4
Degree course 0.4 1.2 2.1 1.9 1.7
n (observations) 1146 954 865 875 844
Panel C: women
Any course 10.7 15.0 18.2 14.3 14.4
Degree course 0.8 1.3 2.2 3.1 3.6
n (observations) 933 741 720 734 724

Higher Education
1 3
undertaking paid employment (OR = 0.68, p < 0.01); and having been in Australia less than
1year (OR = 0.45, p < 0.01) or 1year (OR = 0.74, p < 0.05), compared to 2years.
The magnitude of these associations is better grasped by visual inspection of the left
panel of Fig. 1, in which the model results are transformed into predicted probabili-
ties. This hints at which of the relationships of interest are substantially—as opposed to
statistically—significant. In what follows, we discuss predictors that were both statisti-
cally and substantially significant. As can be appreciated in the figure, the differences in
Table 2 Odds ratios from
random-effect logistic regression
models
BNLA, waves 1–5 (2013–2017)
a Conditional on studying for a qualification of any kind. Statistical sig-
nificance: *p < 0.1, **p < 0.05, ***p < 0.01
Currently studying
Any course Degreea
(1) (2)
Already completed Australian qualification 0.18*** 3.26*
Respondent is a woman 1.54*** 2.62*
Age 1.03 0.85
Age squared 1.00 1.00
Spoken English proficiency (1–4) 2.43*** 4.46***
General health (1–6) 1.12*** 1.33
Level of education pre-arrival (ref. Degree qualification)
None 0.36*** 0.01**
Some schooling 0.46*** 0.09***
Trade qualification 0.63*0.21
Marital status (ref. Married/Partnered)
Divorced, separate or widowed 1.12 2.11
Single, never married 2.43*** 2.92
Number of children in the household 1.01 0.71
Onshore migrant 0.59*** 1.03
Lives in a regional area 1.32 2.50
Currently in paid employment 0.68*** 0.85
Ever in paid employment pre-arrival 0.84 2.67*
Time in Australia (ref. 2years)
< 1year 0.45*** 0.41
1year 0.74** 0.63
3years 0.85 5.58***
4 + years 0.84 8.61***
Country of origin (ref. Iraq)
Afghanistan 1.44** 2.09
Iran 2.48*** 3.14
Myanmar 5.22*** 0.07
Other country 1.76*** 2.30
n (observations) 8492 1070
n (groups) 2109 668
Pseudo R20.04 0.12

Higher Education
1 3
humanitarian migrants’ access to the Australian education system were greatest across
the categories of pre-arrival education, country of birth, English-language proficiency
and—to a lesser extent—general health. Take for example the observed differences
by the educational qualifications humanitarian migrants held at the point of arriving
in Australia. Adjusting for the model controls, those who entered Australia with no
qualifications participated in the education system in 10.1% of the observations. The
analogous figures are visibly higher amongst those who entered Australia with some
schooling (11.9%), a trade qualification (14.5%) and—particularly—a university degree
(19%). The gradient for English-language proficiency is equally striking: respondents
who reported speaking ‘not at all’ English had a 5% probability of being a student,
compared to a much higher 28.5% amongst respondents who reported speaking English
‘very well’.
Predictors ofhigher‑education access.
Column (2) in Table2 shows the results of a model examining which characteristics were
associated with individuals studying towards a degree rather than other courses, out of
Fig. 1 Predicted probabilities from random-effect logistic regression models of studying towards an Aus-
tralian qualification and studying towards a degree. BNLA, waves 1–5 (2013–2017). Based on results from
the model presented in columns (1) and (2) in Table2. For discrete variables (e.g. marital status), we pre-
sent predicted probabilities for each of the variable’s categories. For continuous variables (e.g., age), we
selected representative values of the variable’s distribution and presented predicted probabilities for those
values. Orange and green colouring is used to visually separate results for different, adjacent variables. Bars
without filling denote discrete variables in which none of the categories was statistically significant relative
to the reference category or continuous variables that were not statistically significant

Higher Education
1 3
those individuals who were studying towards any type of qualification.3 The model coef-
ficients indicated that such odds were significantly higher amongst women than men
(OR = 2.45, p < 0.1), and those who had been employed pre-arrival compared to those who
had not (OR = 2.67, p < 0.1), and increased with English-language proficiency (OR = 4.46,
p < 0.01) and time since arrival. The odds of studying towards a degree were, however,
negatively associated with having arrived in Australia with no qualifications (OR = 0.01,
p < 0.05) or only some schooling (OR = 0.09, p < 0.05), compared to having arrived with
a university degree. Of note, some of the ORs in the model appear to be large but are not
statistically significant. This is likely due to reduced statistical power stemming from the
smaller sample size in this model, which is restricted to individuals who were studying for
a course.
A visual representation of the magnitude of these results as predicted probabilities is
displayed in the right panel of Fig.1. Again, here we discuss only predictors that were both
statistically and substantially significant. The figure shows that the most pronounced differ-
ences were for pre-arrival education, English-language proficiency and time since arrival.
This pattern is therefore similar to the one observed for studying towards any qualifica-
tion. However, some of these factors appear particularly important for studying towards
a university degree, including prior educational qualifications and time since arrival. For
instance, differences by educational qualifications pre-arrival indicated that those who
entered Australia with no qualifications selected a HE course in 4.1% of observations,
compared to 9.4% amongst those who entered with some schooling, 13% amongst those
with a trade qualification and 21.3% amongst those with a university degree. For time since
arrival, a substantially larger share of humanitarian migrants opted for HE courses when
they had been in Australia for 3years (16.1%) or 4 + years (18.6%), compared to less than
1year (5.8%), 1year (7%) or 2years (8.5%). Other factors that appear particularly relevant
for HE access include having a younger age (20years or younger) and having no children.
Gender differences
In a final set of analyses, we examined whether there were differences between humanitar-
ian-migrant men and women in the predictors of being a (HE) student.4 The results of these
interactive models—presented in Table7—are quite extensive and difficult to interpret.
For this reason, the discussion here focuses on a selection of key findings; specifically, the
results of socio-demographic factors for which statistically significant gender differences
were observed. The model results revealed statistically significant gender differences for
just a handful of variables: having previously completed an Australian qualification, time
spent in Australia, employment status and marital status. A review of the predicted prob-
abilities associated with these parameters (not shown) indicated that the most substantial
differences were those pertaining to time in Australia, employment status and marital sta-
tus. These are discussed in turn, and displayed in Fig.2 as predicted probabilities.
The top panel of Fig.2 demonstrates that both humanitarian-migrant men and women
appear to engage with the education system at similar rates in the first two years in the
4 Due to small cell sizes when disaggregating the sample by gender, it was not possible to undertake anal-
yses of whether the qualifications that humanitarian migrants studied for or had attained were university
degrees or other type of courses.
3 The results of an alternative model specification comparing individuals studying a degree to all other
individuals, including non-students, were very similar (see Appendix 1 Table6).

Higher Education
1 3
country. However, while women’s rates grow over time, men’s rates decrease. This may
reflect a propensity for humanitarian-migrant women to take the bulk of the domestic
responsibilities over the early settlement period (see, e.g. Hatoss & Huijser, 2010; Seck,
2015). The middle panel of Fig.2 shows how paid employment seems to be somewhat of a
barrier to accessing education amongst humanitarian-migrant men, but markedly increases
the probability of accessing the education system amongst humanitarian-migrant women.
Fig. 2 Predicted probabilities
from random-effect logistic
regression models of the odds
of being a student, by gender
and time in Australia, employ-
ment status and marital status.
BNLA, waves 1–5 (2013–2017).
Based on results from the model
presented in Table7

Higher Education
1 3
It is possible that men and women are opting for different types of qualifications, with dif-
ferent purposes: women may target qualifications to upskill while already in a job, while
men may target qualifications that improve their chances of moving quickly into employ-
ment. Either way, this pattern of results indicates that a greater share of humanitarian-
migrant women than men juggle work and study. Finally, the bottom panel of Fig.2 makes
it apparent that being single is a significant incentive for humanitarian-migrant women but
not men to become a student. This may reflect a reliance on a male-breadwinner model
amongst humanitarian-migrant women, with those who are single being incentivised to
upskill and, in doing so, augment their labour-market prospects.
Discussion andconclusion
This study set out to contribute unique empirical evidence on the socio-demographic fac-
tors that structure inequalities in humanitarian migrants’ access to education, and HE more
specifically, using a unique quantitative dataset (BNLA). The analyses yielded several key
findings. First, while only a small share of humanitarian migrants started an educational
course early into their settlement period, an upwards trend over the 5-year observation
window was observed. By the end of the fifth year, 15.4% of humanitarian migrants were
enrolled in a course (2.6% in a HE course). Although this time trend is a reason for mod-
erate optimism, engagement with the HE system amongst humanitarian migrants in Aus-
tralia is modest: of all students, only 14.2% pursued HE. This evidence supports the notion
that this population group does indeed experience inequalities accessing the Australian HE
system.
Second, our results provide strong evidence of within-group inequalities in access to
HE amongst humanitarian migrants in Australia. Two factors consistently predicted greater
engagement with the Australian education system and with higher-order options within it:
(i) English-language proficiency and (ii) education level at the time of arrival. The impor-
tance of other factors (e.g. age, general health, marital status, parenthood and onshore
migration) was smaller. The critical role of mastery in the English language observed
here is, however, not surprising. In fact, improving local-language proficiency has been
identified by the OECD as a priority area for the educational integration of humanitar-
ian migrants (Cerna, 2019, p.34). Cerna (2019) discusses several strategies across OECD
countries aimed at improving humanitarian migrants’ skills in the host-country language,
including introductory/welcome language courses and transition-to-mainstream-lan-
guage programs. In Australia, the Adult Migrant English Program (AMEP) has histori-
cally provided up to 510h of English-language tuition to eligible humanitarian migrants
to help them learn the foundations of the English language, as well as cultural skills that
enable them to participate socially and economically in Australian society. Our findings
suggest that increasing humanitarian migrants’ access to this type of programs and/or
the intensity of their exposure may go a long way in enhancing their educational pros-
pects. In this regard, in August 2020, the Australian government made positive changes
to the AMEP, including removing the cap of 510h and the time limits for enrolment and

Higher Education
1 3
completion—which previously meant that humanitarian migrants had to complete the pro-
gram within 5years of settlement in Australia (Tudge, 2020).
Similarly, arriving in Australia with degree-level education qualifications was also a fac-
tor that consistently predicted engagement in the Australian education system, including
HE. This finding is somewhat concerning, as it suggests that those humanitarian migrants
with greater need to access the Australian education system (i.e. those with low or no edu-
cational credentials at arrival) are those who appear to be least likely to access it. And
this finding does not simply reflect that highly-skilled humanitarian-migrants may be more
likely to have sufficient educational credentials to be admitted into a HE course. In fact,
this group was also more likely than the low-skilled group to access non-HE courses. A
possible explanation is that highly educated humanitarian migrants may arrive in Australia
with advanced skills and an appreciation of the importance of education for improving
their socio-economic prospects, but may hold education credentials that are not officially
recognised. For this group of humanitarian migrants, access to education (particularly HE)
may be a pathway to validate their existing skills with a formal Australian qualification.
Altogether, these results underscore the importance of reaching low-skilled humanitarian
migrants for any policies or programs aimed at improving humanitarian-migrants’ educa-
tional trajectories in Australia. Policies aimed at improving the educational engagement of
low-skilled humanitarian migrants may focus on developing their HE aspirations, improv-
ing their financial capacity to enrol and developing the skills required for successful par-
ticipation (e.g. general literacy and numeracy).
Third, the results revealed significant differences in the probability of studying a course
by humanitarian-migrants’ country of origin. For example, humanitarian migrants from
Iraq were generally found to display lower engagement with education than humanitar-
ian migrants coming from other major source countries. It is possible that humanitarian
migrants from Iraq experience the most elevated rates of trauma owing to the intensity,
duration and contemporaneity of conflict in their country. Indeed, a wealth of psychological
literature documents the long-term consequences of trauma amongst humanitarian-migrant
populations post resettlement (Slewa-Younan et al., 2015). The latter may also prompt
Iraqi humanitarian migrants in Australia to prioritise employment over study options, as
a means to send remittances to their family members back in their country. Either way,
this finding echoes previous research highlighting the diversity in humanitarian-migrant
experiences (e.g. Cerna, 2019; Naylor etal., 2019) and the ensuing complexity in design-
ing intervention programs (OECD, 2016). The observed pattern of results is clearly a cause
for concern, as Iraqi entrants are the largest group of recent humanitarian migrants in Aus-
tralia—amounting to 41.5% of those who arrived in the 2018/2019 period (DHA, 2019). It
is nevertheless important to acknowledge that we cannot separately examine the outcomes
of humanitarian migrants from small source countries, and it is possible that humanitar-
ian migrants from some of these countries experience more disadvantageous circumstances
than Iraqi humanitarian migrants. Regardless, this finding underscores the importance of
considering specific countries of origin when providing targeted support to humanitarian
migrants.
Inspired by scholarship documenting important gender disparities in the socio-
economic integration of immigrant men and women, an additional set of analyses

Higher Education
1 3
examined the potential gendering of the processes discussed thus far. Specifically,
the analyses tested for gender differences in the factors enabling/constraining access
to education within this population. A core finding emerging from these analyses was
that—compared to humanitarian-migrant men in Australia—humanitarian-migrant
women exhibited a greater adjusted likelihood of being a student. The timing of edu-
cation access differed somewhat between humanitarian migrants of either gender: the
share of humanitarian-migrant men in education decreased with time since arrival,
whereas the share of humanitarian-migrant women in education increased with
time since arrival. The absence of a ‘female penalty’ in relation to education access
amongst humanitarian migrants in Australia constitutes an interesting finding. This is
because, compared to humanitarian-migrant men, humanitarian-migrant women are
disadvantaged in other life domains, most notably in relation to their labour-market
outcomes (ABS, 2018; Delaporte & Piracha, 2018; OECD, 2019). Similarly, while
it is theoretically plausible that humanitarian-migrant women experienced different
challenges to accessing education than humanitarian-migrant men, the BNLA results
portrayed a picture of similarity rather than one of difference. There were few statisti-
cally significant gender differences in the estimated effect of the socio-demographic
predictors on the likelihood of studying.
Despite the importance of our findings, there are also study limitations that must
be acknowledged, some of which point to potentially fruitful avenues for further
research. First, the humanitarian migrants covered in our data exclude refugees on
temporary protection visas and safe haven enterprise visas, who are arguably the
most disadvantaged subgroups in relation to accessing the Australian (higher) edu-
cation system (Hartley etal., 2018). Second, while BNLA offers important advan-
tages in terms of the richness of the data collected and its longitudinal design, it
does not enable direct comparisons between humanitarian migrants and other popu-
lation groups. Third, the data pertain to a single cohort of humanitarian migrants
who arrived at a specific socio-historical point in time and whose experiences and
circumstances may or may not coincide with those of newer cohorts with different
characteristics and pre-settlement pathways. Finally, our analyses are not designed
to identify causal relationships; future research deploying causal modelling and/
or more direct, self-reported measures of ‘barriers to education’ would be better
placed to claim causality.
Despite these limitations, the findings presented here carry important lessons for
policy and practice. They reveal that specific segments of the humanitarian-migrant
population are comparatively less likely to access Australian HE. Hence, they sug-
gest that a ‘one size fits all’ policy strategy may be neither sufficient nor appropriate
to boost the chances of humanitarian migrants reaping the benefits of HE in the host
society. Instead, certain subpopulations within the broader humanitarian-migrant
group require special attention from equity practitioners and policymakers—for
example, those with low levels of English-language proficiency and those entering
the country with low/no educational qualifications.

Higher Education
1 3
Appendix 1: Additional tables
Tables3, 4, 5, 6, and7
Table 3 Sample means and
standard deviations for analytic
variables
BNLA, waves 1–5 (2013–2017). SD standard deviation
Mean/% SD
Currently studying in Australia (%)
Any course 12.6
Degree course 1.7
Trade/technical course, or a paid traineeship 5.0
Some other course, or work experience 5.3
Woman (%) 45.3
Age 37.3 11.6
Spoken English proficiency (1–4) 2.3 0.8
General health (1–6) 3.9 1.4
Level of education pre-arrival (%)
None 15.2
Some schooling 67.8
Trade qualification 6.1
Degree qualification 10.8
Marital status (%)
Married, partnered 62.6
Divorced, separate or widowed 10.0
Single, never married 27.4
Number of children in the household 1.1 1.2
Onshore migrant (%) 14.7
Lives in a regional area (%) 8.5
Currently in paid employment (%) 20.6
Ever in paid employment pre-arrival (%) 56.0
Time in Australia (%)
< 1year 21.0
1year 20.1
2years 18.8
3years 19.1
4 + years 21.0
Country of origin (%)
Iraq 41.7
Afghanistan 25.6
Iran 11.1
Myanmar 5.8
Other country 1.1
n (observations) 8668
n (groups) 2109

Higher Education
1 3
Table 4 Sample means for analytic variables using different subsamples
BNLA, waves 1–5 (2013–2017). SD standard deviation
All respondents Only students Only
university
students
Currently studying in Australia (%)
Any course 12.6 100.0 100.0
Degree course 1.7 13.6 100.0
Trade/technical course, or a paid traineeship 5.0 39.8 4.1
Some other course, or work experience 5.3 42.1 4.1
Woman (%) 45.3 51.4 56.2
Age 37.3 32.8 29.0
Spoken English proficiency (1–4) 2.3 2.7 3.4
General health (1–6) 3.9 4.3 4.8
Level of education pre-arrival (%)
None 15.2 7.9 0.7
Some schooling 67.8 68.2 58.9
Trade qualification 6.1 5.9 3.4
Degree qualification 10.8 17.9 37.0
Marital status (%)
Married, partnered 62.6 45.7 28.1
Divorced, separate or widowed 10.0 7.5 4.8
Single, never married 27.4 46.8 67.1
Number of children in the household 1.1 1.0 0.6
Onshore migrant (%) 14.7 14.6 22.6
Lives in a regional area (%) 8.5 10.7 12.3
Currently in paid employment (%) 20.6 20.5 26.7
Ever in paid employment pre-arrival (%) 56.0 51.3 59.6
Time in Australia (%)
< 1year 21.0 17.7 6.2
1year 20.1 21.6 11.6
2years 18.8 21.9 17.8
3years 19.1 19.5 30.1
4 + years 21.0 19.3 34.2
Country of origin (%)
Iraq 41.7 34.5 32.9
Afghanistan 25.6 19.1 12.3
Iran 11.1 18.5 29.5
Myanmar 5.8 9.1 0.7
Other country 1.1 18.9 24.7
n (observations) 8668 1070 146
n (groups) 2109 668 89

Higher Education
1 3
Table 5 Odds ratios from random-effect logistic regression models, alternative age ranges
BNLA, waves 1–5 (2013–2017). Statistical significance: *p < 0.1, **p < 0.05, ***p < 0.01
Currently studying: Any course
Age range: 18 to 51years 18 to 31years
Already completed Australian qual 1.05 1.08
Respondent is a woman 1.00 1.00
Age 0.18*** 0.19***
Age squared 1.61*** 2.00***
Spoken English proficiency (1–4) 2.46*** 2.48***
General health (1–6) 1.07*1.07
Education pre-arrival (ref. Degree)
None 0.34*** 0.16***
Some schooling 0.45*** 0.36***
Trade qualification 0.53** 0.32**
Marital status (ref. Married/Partnered)
Divorced, separate or widowed 1.03 1.33
Single, never married 2.55*** 3.46***
Number of children in household 1.01 0.97
Onshore migrant 0.56*** 0.43***
Lives in a regional area 1.17 1.41
Currently in paid employment 0.65*** 0.60***
Ever in paid employment pre-arrival 0.92 0.97
Time in Australia (ref. 2years)
< 1year 0.39*** 0.22***
1year 0.73** 0.64**
3years 0.86 1.02
4 + years 0.85 1.24
Country of origin (ref. Iraq)
Afghanistan 1.44** 1.01
Iran 2.26*** 1.41
Myanmar 4.50*** 3.42***
Other country 1.71*** 1.47
n (observations) 7171 2882
n (groups) 1828 839
Pseudo R20.04 0.04

Higher Education
1 3
Table 6 Odds ratios from a
random-effect logistic regression
model of studying towards a
university degree, assigning the
value zero to non-students
BNLA, waves 1–5 (2013–2017). Statistical significance: *p < 0.1,
**p < 0.05, ***p < 0.01
Currently
studying:
degree
Already completed Australian qual 0.39***
Respondent is a woman 3.13***
Age 0.89
Age squared 1.00
Spoken English proficiency (1–4) 4.68***
General health (1–6) 1.29**
Education pre-arrival (ref. Degree)
None 0.02***
Some schooling 0.15***
Trade qualification 0.22**
Marital status (ref. Married/Partnered)
Divorced, separate or widowed 2.28
Single, never married 4.19***
Number of children in household 0.78
Onshore migrant 0.59
Lives in a regional area 2.20
Currently in paid employment 0.60
Ever in paid employment pre-arrival 1.75
Time in Australia (ref. 2years)
< 1year 0.14***
1year 0.37**
3years 2.37**
4 + years 2.87***
Country of origin (ref. Iraq)
Afghanistan 1.28
Iran 3.15**
Myanmar 0.31
Other country 1.83
n (observations) 8492
n (groups) 2109
Pseudo R20.10

Higher Education
1 3
Table 7 Odds ratios from a
random-effect logistic regression
models with gender interactions
Currently
studying: any
course
Main effects
Woman 0.21
Already completed Australian qualification 0.24***
Age 0.99
Age squared 1.00
Spoken English proficiency (1–4) 2.29***
General health (1–6) 1.11**
Education: Degree (ref.)
Education: None 0.42**
Education: Some schooling 0.60**
Education: Trade qualification 0.72
Marital status: Married/Partnered (ref.)
Marital status: Divorced, separate or widowed 1.80
Marital status: Single, never married 1.40
Number of children in the household 1.04
Onshore migrant 0.61**
Lives in a regional area 1.62**
Currently in paid employment 0.55***
Ever in paid employment pre-arrival 0.73*
Time in Australia: 2years (ref.)
Time in Australia: < 1year 0.52***
Time in Australia: 1year 0.93
Time in Australia: 3years 0.88
Time in Australia: 4 + years 0.74
Country of origin: Iraq (ref.)
Country of origin: Afghanistan 1.35
Country of origin: Iran 2.84***
Country of origin: Myanmar 6.27***
Country of origin: Other 1.77**
Interaction effects
Woman interacted with…
Already completed Australian qualification 0.45***
Age 1.11
Age squared 1.00
Spoken English proficiency (1–4) 1.12
General health (1–6) 1.01
Education: None 0.95
Education: Some schooling 0.69
Education: Trade qualification 0.85
Marital status: Divorced, separate or widowed 0.71
Marital status: Single, never married 3.15***
Number of children in the household 0.91
Onshore migrant 1.44
Lives in a regional area 0.59

Higher Education
1 3
Appendix 2: Formal model specication
Formally, the random-effect logistic models that we fit take the following form:
where subscripts i and t stand for individual and time period, respectively; E is a
binary measure capturing being enrolled in an education or HE course in Australia
measured at time t; S represents the set of socio-demographic variables described before
and measured at time t; α is a model intercept; β is the vector of coefficients of inter-
est to be estimated; and u is a person-specific random intercept (i.e., a random effect).
While an array of longitudinal approaches are available to analyse panel data of the
sort available in BNLA, this random-effect approach is fit-for-purpose to achieve our
research aim of identifying the socio-demographic correlates of HE access amongst
humanitarian migrants. These random-effect models are more efficient and less data-
demanding than other potential approaches. For example, random-effect models using
lagged explanatory variables would reduce the number of waves available for analysis,
substantially diminishing the sample size. Similarly, within-group fixed-effect models
would preclude consideration of time-constant predictors (e.g. pre-arrival education,
gender, and country of origin), and would be highly inefficient in estimating the coef-
ficients on rarely changing variables (e.g. English-language proficiency, marital status,
or general health).
Acknowledgements This research was supported by the National Centre for Student Equity in Higher
Education (NCSEHE) through its 2019 Research Grants Program and by the Australian Research Council
Centre of Excellence for Children and Families over the Life Course (project number CE140100027). The
analyses use data from Building a New Life in Australia: The Longitudinal Study of Humanitarian Migrants
(BNLA). The BNLA project is a collaborative effort between the Australian Government Department
(1)
log(pr
(
E
it =1)
1−
pr
(
Eit
=1)
)
=𝛼+𝜷Sit +u
i
Table 7 (continued) Currently
studying: any
course
Currently in paid employment 2.86***
Ever in paid employment pre-arrival 1.18
Time in Australia: < 1year 0.78
Time in Australia: 1year 0.61**
Time in Australia: 3years 0.96
Time in Australia: 4 + years 1.38
Country of origin: Afghanistan 1.08
Country of origin: Iran 0.66
Country of origin: Myanmar 0.61
Country of origin: Other 0.95
n (observations) 8492
n (groups) 2109
Pseudo R20.04
BNLA, waves 1–5 (2013–2017). Statistical significance: *p < 0.1,
**p < 0.05, ***p < 0.01

Higher Education
1 3
of Social Services (DSS), the Australian Institute of Family Studies (AIFS) and Colmar Brunton Social
Research.
Funding This research was supported by the National Centre for Student Equity in Higher Education
(NCSEHE) through its 2019 Research Grants Program and by the Australian Research Council Centre of
Excellence for Children and Families over the Life Course (project number CE140100027).
Data availability The data used in this study, Building a New Life in Australia: The Longitudinal Study of
Humanitarian Migrants, can be obtained from the Australian Government Department of Social Services
(for details, see https:// datav erse. ada. edu. au/ datav erse. xhtml? alias= bnla).
Code availability Syntax code to replicate the analyses presented in this paper is available from the corre-
sponding author upon request.
Declarations
Conflict of interest The authors declare no conflict of interest.
References
Australian Bureau of Statistics (ABS). (2017). Education and work, Australia, May 2017. Canberra: Aus-
tralian Bureau of Statistics.
Australian Bureau of Statistics (ABS). (2018). Understanding migrant outcomes—Insights from the Aus-
tralian Census and Migrants Integrated Dataset, Australia, 2016, CAT No. 3417.0. Canberra: Austral-
ian Bureau of Statistics. Available online at: https:// www. abs. gov. au/ AUSST ATS/ abs@. nsf/ Lates tprod
ucts/ 3417. 0Main% 20Fea tures 22016? opend ocume nt& tabna me= Summa ry& prodno= 3417. 0& issue=
2016& num= & view=. Accessed 29 Sept 2021.
Australian Bureau of Statistics (ABS). (2020). Characteristics of recent migrants (November 2019). Availa-
ble online at: https:// www. abs. gov. au/ stati stics/ people/ people- and- commu nities/ chara cteri stics- recent-
migra nts/ latest- relea se. Accessed 29 Sept 2021.
Australian Institute of Family Studies (AIFS). (2018). Building a new life in Australia data users guide
release 4.0. Victoria: Australian Institute of Family Studies. Available online at: http:// www3. aifs. gov.
au/ bnla/ userg uide/ docs/ BNLA% 20Data% 20Use rs% 20Gui de% 20Rel ease% 204. pdf. Accessed 29 Sept
2021.
Bajwa, J. K., Couto, S., Kidd, S., Markoulakis, R., Abai, M., & McKenzie, K. (2017). Refugees higher edu-
cation and informational barriers. Refuge: Canada’s Journal on Refugees, 33(2), 56–65.
Baker, S., Irwin, E., Taiwo, M., Singh, S., Gower, S., & Dantas, J. (2019). Methodological diversity as
an asset for transition-focused higher education research with students from refugee backgrounds.
Review of Education, 7(1), 5–32.
Baker, S., Ramsay, G., & Lenette, C. (2019). Students from refugee and asylum seeker backgrounds and
meaningful participation in higher education: From peripheral to fundamental concern. Widening
Participation and Lifelong Learning, 21(2), 4–19.
Cerna, L. (2019). Refugee education: Integration models and practices in OECD countries. OECD.
Correa-Velez, I., Barnett, A. G., & Gifford, S. (2015). Working for a better life: Longitudinal evidence
on the predictors of employment among recently arrived refugee migrant men living in Australia.
International Migration, 53(2), 321–337.
Delaporte, I., & Piracha, M. (2018). Integration of humanitarian migrants into the host country labour
market: Evidence from Australia. Journal of Ethnic and Migration Studies, 44(15), 2480–2505.
Department of Home Affairs (DHA). (2019). Australia’s Offshore Humanitarian Program: 2018–19.
Canberra, ACT: Commonwealth of Australia.
Desjardins, R., & Lee, J. (2016). Earnings and employment benefits of adult higher education in com-
parative perspective: Evidence based on the OECD Survey of Adult Skills (PIAAC). Los Angeles:
UCLA.
Department of Social Services (DSS). (2017). Building a new life in Australia (BNLA): The longitudi-
nal study of humanitarian migrants—Findings from the first three waves. Canberra: Department of
Social Services.

Higher Education
1 3
Earnest, J., Joyce, A., De Mori, G., & Silvagni, G. (2010). Are universities responding to the needs of
students from refugee backgrounds? Australian Journal of Education, 54(2), 155–174.
Harris, V., & Marlowe, J. (2011). Hard yards high hopes: The educational challenges of African refugee
university students in Australia. International Journal of Teaching, Learning and Higher Educa-
tion, 23(2), 186–196.
Harris, A., Spark, C., & Ngum Chi Watts, M. C. (2015). Gains and losses: African Australian women
and higher education. Journal of Sociology, 51(2), 370–384.
Hartley, L., Fleay, C., Baker, S., Burke, R., & Field, R. (2018). People seeking asylum in Australia:
Access & support in higher education. National Centre for Student Equity in Higher Education,
Curtin University.
Hatoss, A., & Huijser, H. (2010). Gendered barriers to educational opportunities: Resettlement of Suda-
nese refugees in Australia. Gender Education, 22(2), 147–160.
Hugo, G. (2011). A significant contribution: the economic, social and civic contributions of first and sec-
ond generation humanitarian entrants: Summary of findings. Commonwealth of Australia, DIAC.
Joyce, A., Earnest, J., De Mori, G., & Silvagni, G. (2010). The experiences of students from refugee
backgrounds at universities in Australia: Reflections on the social, emotional and practical chal-
lenges. Journal of Refugee Studies, 23(1), 82–97.
Krieger, M. (2020). Tied and troubled: Revisiting tied migration and subsequent employment. Journal of
Marriage and Family, 82(3), 934–952.
Mangan, D., & Winter, L. A. (2017). (In)validation and (mis)recognition in higher education: The expe-
riences of students from refugee backgrounds. International Journal of Lifelong Education, 36(4),
486–502.
Manhica, H., Berg, L., Almquist, Y. B., Rostila, M., & Hjern, A. (2019). Labour market participation
among young refugees in Sweden and the potential of education: A national cohort study. Journal
of Youth Studies, 22(4), 533–550.
Morrice, L. (2009). Journeys into higher education: The case of refugees in the UK. Teaching in Higher
Education, 14(6), 661–672.
Naidoo, L. (2019). Traversing the terrain of higher education: Experiences of refugee youth on the
inside. International Journal of Inclusive Education, 1–14.
Naidoo, L., Wilkinson, J., Adoniou, M., & Langat, K. (2018). Refugee background students transitioning
into higher education: Navigating complex spaces. Springer.
Naylor, R., Terry, L., Rizzo, A., Nguyen, N., & Mifsud, N. (2019). Structural inequality in refugee par-
ticipation in higher education. Journal of Refugee Studies, fez077.
Organisation for Economic Co-operation and Development (OECD). (2016). Making integration work:
Refugees and others in need of protection. OECD Publishing.
Organisation for Economic Co-operation and Development (OECD). (2019). International migration
outlook 2019. OECD Publishing.
Organisation for Economic Co-operation and Development (OECD) & European Union (2018). Settling
In 2018: Indicators of immigrant integration (Chapter6). OECD Publishing, Paris/European Union,
Brussels. Available online at: https:// www. oecd- ilibr ary. org/ docse rver/ 97892 64307 216- 10- en. pdf?
expir es= 16328 24551 & id= id& accna me= ocid1 77546 & check sum= C83CE F0AF8 69371 D1D74
82F11 F7DCE 96. Accessed 29 Sept 2021.
Ramsay, G., & Baker, S. (2019). Higher education and students from refugee backgrounds: A meta-scoping
study. Refugee Survey Quarterly, 38(1), 55–82.
Seck, M. M. (2015). Female West African immigrants in the United States: Challenges and empowering
strategies. Social Development Issues, 37(2), 68–79.
Slewa-Younan, S., Mond, J. M., Bussion, E., Melkonian, M., Mohammad, Y., Dover, H., . . . Jorm, A. F.
(2015). Psychological trauma and help seeking behaviour amongst resettled Iraqi refugees in attending
English tuition classes in Australia. International Journal Mental Health Systems, 9(5). https:// doi. org/
10. 1186/ 1752- 4458-9-5
Stevenson, J., & Baker, S. (2018). Refugees in higher education: Debate, discourse and practice: Emerald
Publishing Limited.
Streitwieser, B., Loo, B., Ohorodnik, M., & Jeong, J. (2018). Access to higher education for refugees: A
review of interventions in North America and Europe. Graduate School of Education & Human Devel-
opment, the George Washington University.
Summerfield, M., Garrard, B., Hahn, M., Jin, Y., Kamath, R., Macalalad, N., Watson, N., Wilkins, R., &
Wooden, M. (2020). HILDA User Manual – Release 19. Applied Economic and Social Research, Uni-
versity of Melbourne.
Szkudlarek, B. (2019). Engaging business in refugee employment: The employers’ perspective. Sydney: the
University of Sydney.

Higher Education
1 3
Terry, L., Naylor, R., Nguyen, N., & Rizzo, A. (2016). Not there yet: An investigation into the access and
participation of students from humanitarian refugee backgrounds in the Australian Higher Education
system. Perth: National Centre for Student Equity in Higher Education.
Tudge, A. (2020). More class hours available to help migrants learn English. Available online at: https://
minis ter. homea ffairs. gov. au/ alant udge/ Pages/ More- class- hours- avail able- to- help- migra nts- learn- Engli
sh. aspx. Accessed 29 Sept 2021.
United Nations High Commissioner for Refugees (UNHCR). (2019). Refugee education 2030: A strategy
for refugee inclusion: United Nations High Commissioner for Refugees.
Watkins, P. G., Razee, H., & Richters, J. (2012). ‘I’m Telling You … The language barrier is the most, the
biggest challenge’: Barriers to education among Karen refugee women in Australia. Australian Journal
of Education., 56(2), 126–141. https:// doi. org/ 10. 1177/ 00049 44112 05600 203
Wilkins, R., & Lass, I. (2018). Household, income and labour dynamics in Australia survey: Selected find-
ings from waves 1 to 16: the 13th Annual Statistical Report of the HILDA Survey. Melbourne: Mel-
bourne Institute of Applied Economic and Social Research.
Wooldridge, J. M. (2010). Econometric analysis of cross section and panel Data. MIT Press.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.