Population Change and Lifecourse

Research Brief #21 - October 2014

Academic Performance and Educational Pathways of Allophone Youth: A Comparative Analysis of Montreal, Toronto, and Vancouver


This study examines the academic performance and educational pathways of students who do not speak the language of schooling at home—that is, French in Montreal and English in Toronto and Vancouver. Overall, we discover that these students, who consist of almost exclusively allophones, graduate more or less as other students but, when their personal characteristics are controlled, they appear to succeed much better, especially in Vancouver. However, there are substantial differences in performance between linguistic subgroups of allophones. Further, these subgroups tend to show varied results by city. Educational authorities should therefore pay special attention to the criteria used to allocate funds to schools with a high concentration of these groups, as “one-size fits all” support is not evidence-based.

Key Findings


Qualitative research has shown that first- and second-generation youth in Canada tend to face difficulties with academic integration, particularly if they do not have French or English as their mother tongue and belong to visible minorities (Anisef and Kilbride 2001). However, there have not been enough large-scale,comparative quantitative studies to assess the academic performance and educational pathways of immigrant origin students. This paucity can be explained by the fact that education falls under different provincial and territorial authorities, each with its own educational structures, policies, and ways of collecting educational data. Our study contributes to filling this gap in research.

We also examine the factors that could explain school achievement among immigrant origin youth. Prior research has focused on many different variables. For instance, socioeconomic theories state that poverty is generally associated with a lack of involvement in educational promotion from parents; therefore, poorer immigrant youth tend to experience lower achievement (Portes and Zhou 1993). Other researchers have argued that reactions of school systems to immigrants are more important, while some have focused on characteristics of home country culture and how these are linked to success in Canadian school systems (Gillborn and Gipps 1996, Peng and Wright 1994). Although we cannot examine all potential factors in our study, we do look at a number of socio-demographic and contextual variables, such as gender, median income, and attendance at private school, to evaluate which of these have consistently significant effects on the academic successes of immigrant origin youth.

Data and Method

This study uses provincial or local administrative data from Vancouver, Montreal, and Toronto, the three major immigrant-destination cities in Canada. In each site, it follows the school pathways of the full cohort of students who were expected to graduate in 2004—that is, who entered the first level of secondary school in 1999 in Montreal and Toronto, and in 2000 in Vancouver.

Initially, we aimed to identify youth with an immigrant origin as our group of study. Unfortunately, immigration status and ethnicity data were not available for students in our study sites. Instead, we use home language data. Specifically, our target group consists of students in the educational system who use a language at home other than the language of schooling. For lack of a better word, we use the term “allophones” to designate this group. It largely consists of first-and second-generation students who still use the language of their parents. The comparison group includes students who speak the language of schooling at home: French in Montreal and English in Toronto and Vancouver.

Our study specifically examines academic attainment (graduation) and educational pathways (participation in a course leading to college/university) of the target group, contrasting these results with those of the comparison group. We investigated (1) whether the target group fares better than the comparison group, and (2) what factors significantly influence their academic attainment and pathways. It is important to note that although we can contrast trends, there are too many differences between the three contexts here to permit a strict comparison of academic successes of our groups. Indeed, there are significant variations in school system structures across these sites, as well as in the variables found in data banks. Nevertheless, our results allow for general conclusions and comments.


Descriptive Analysis

In all three sites, we discovered that students in the target group graduate at similar rates to those in the comparison group, with allophone rates being slightly lesser in Montreal and slightly higher in Vancouver (see Table 1). Notably, however, there are differences in rates by linguistic subgroup. For instance, the rate of graduating Chinese-speaking students was 78.1% in Toronto, as compared to 64.8% for English speakers. However, for Spanish speakers, the rate was 46.9%. Several linguistic subgroups also graduate at varied rates by city. The Arabic subgroup, for example, has a significantly higher rate of graduation in Vancouver and Montreal than it does in Toronto.

Research Brief 21 Table 1

Note: Tables 1 and 2 do not include results from all linguistic subgroups. See the journal article for more details.

Data regarding participation in a selective or university-bound course reveal that such participation is higher in the target group than in the comparison group for all three cities. This result is especially noticeable in Vancouver, where the target group participation rate is 54%, as compared to 28% for the comparison group. As shown in Table 1, there are also wide variations in participation rates for each linguistic subgroup. Overall, in all three sites, those in the target group tend to do better than the comparison group with respect to participation, rather than graduation.

Multivariate Analysis

We shift our study to include more variables, both to control for the characteristics of our subgroups, and to assess the impact of these characteristics on academic achievement. We used the following ten control variables,sorted here by category:

Likelihood of Graduation

Using regression analysis with the above control variables, we found that the target group succeeds better in terms of graduation than the comparison group, especially in Vancouver (see Table 2). Allophones are 2.12 times more likely than English-speakers to graduate secondary school in Vancouver, and are 1.35 times more likely in Toronto. In Montreal, they are 1.39 times more likely to graduate than French-speaking students.

Research Brief 21 Table 2

In terms of linguistic subgroups, Chinese speakers have much higher odds of graduating than the other subgroups at all three sites. Further, the Vietnamese subgroup shows very different odds in Montreal and Vancouver. Inter-city differences could confirm the impact of different family and community values, positive/negative relationships with the host society, or systemic factors such as the teacher’s attitude or valorization of specific languages and cultures.

Our analysis also indicates that four variables significantly affect graduation of the target group in all three sites (see full article for this data table). Being a female and attending a private school both have a positive effect on graduation for the target group, while arriving late to secondary school and changing schools have a negative effect. The rest of the variables do not have a significant influence on graduation in all cities. For instance, median family income matters in Toronto, but not in Montreal or Vancouver. These results show that certain characteristics are not clear predictors for the educational experiences of immigrant origin youth across Canada.

Likelihood of Participation in a Selective or University-Bound Course

Data indicate that our target group is much more likely to participate in a selective or university-bound course than the comparison group. Again, this is especially true for Vancouver, where allophones are 2.73 times more likely to participate in such courses than other students, versus 1.82 times in Toronto and 1.49 times in Montreal (see Table 2). Furthermore, there are again large differences between linguistic subgroups: for instance, the Chinese subgroup shows relatively high odds of participation, while the Spanish subgroup has much lower odds.

As with graduation, we found that the variables age of arrival, changing schools, and private school attendance similarly affect the target group’s participation in these courses. However, income is a much more significant factor for participation than it is for graduation. Further, allophone students entering the school system in the first year of secondary school, as opposed to entering during primary school, are more likely to continue on to university. Finally, as mentioned previously, being a female increases chances of graduating; however, it reduces the likelihood of participating in a selective or university-bound course. See the full article for more details on these variables.


Firstly, there are several factors that consistently influence academic success for allophones. Age on arrival, school change, gender, and attendance at a private school are all significant to their graduation and educational pathways. For future study, we would like to use alternative strategies to further examine how socioeconomic status of immigrant families affects their children’s academic success. We would also like to examine various schooling factors in more depth. For instance, it would be interesting to investigate why private schools are doing a much better job with allophone students than public schools.

Secondly, linguistic subgroups are very important when examining academic performance. Although the target group appears to succeed better than the comparison group, subgroups show very substantial differences, with Chinese speakers appearing consistently on top and Spanish speakers on the bottom. This result is congruent with findings in recent literature. Although a more in-depth look at subgroup differences is necessary, current findings should generate an understanding for educational authorities on the way they organize their services and programs for immigrant origin students. Special attention should be paid to criteria used to allocate special funds to schools with a high concentration of our target group. Considering that allophone subgroups can be four times more or less likely to graduate than the comparison group, it is obvious that “one size fits all” support is not an evidence-based policy.



• Anisef, P., and K.M. Kilbride. 2001. To build on hope: Overcoming the challenges facing newcomer youth at risk in Ontario. Toronto: CERIS.
• Gillborn, D., and C. Gipps. 1996. Recent Research on the Achievements of Ethnic Minority Pupils. London: Office for Standards in Education.
• Peng, S.S., and D. Wright. 1994. Explanation of academic achievement of Asian American students. Journal of Educational Research 87(6):346–352.
• Portes, A., and M. Zhou. 1993. The new second generation: Segmented assimilation and its variants. Annals of the American Academy of Political and Social Science 530:74–96.

About the Study

This research brief is based on Jacques Ledent et al, "Academic performance and educational pathways of young allophones: A comparative multivariate analysis of Montreal, Toronto, and Vancouver”, Canadian Studies in Population 40, no. 1-2(2013): 35-56.

For further information please contact Jacques Ledent, Institut national de la recherche scientifique

This research brief was written by Carmina Ravanera.