Regional differences in the educational outcomes of young immigrants
by Feng Hou and Qi Zhang
- Overview of the study
- PISA scores in reading and math vary across regions
- Explaining differences in test scores between immigrants and third- or higher-generation students
- Why do immigrants have better PISA scores in some regions than others?
- Regional differences in high-school and university completion rates
- Related material for this article
Start of text box
This article examines regional differences in the math and reading skills of immigrant children aged 15 based on data from the Programme for International Student Assessment (PISA). It also examines regional differences in high-school and university completion rates among young immigrants who came to Canada before the age of 15 using National Household Survey (NHS) data. Throughout the article, comparisons are made with the children of the Canadian-born (third- or higher-generation Canadians).
- In Canada, the average PISA math score of immigrant students aged 15 was similar to the score of third- or higher-generation students. The average PISA reading score of immigrant children was slightly lower than the score of third- or higher-generation children.
- In almost all regions, immigrant students had lower PISA reading scores than third- or higher-generation students. With respect to PISA math scores, immigrant students performed better than third- or higher-generation students in the Atlantic provinces and British Colombia, but performed less well in Quebec and in Manitoba and Saskatchewan.
- Young immigrants aged 20 to 24 were more likely to have a high school diploma than their third- or higher-generation counterparts (93% versus 87%). Young immigrants aged 25 to 29 were also more likely to have a university degree (40%, compared with 26% of third- or higher-generation individuals in this age group).
- Manitoba and Saskatchewan (29%) and Quebec (32%) had the lowest proportions of immigrants aged 25 to 29 with a university degree. In contrast, British Columbia (44%) and Ontario (41%) had the highest proportions.
- Regional differences in the source countries of immigrants explained, in part, why some regions had higher university completion rates than others.
End of text box
Canada is well-known for its success in integrating the children of immigrants into the education system and labour market. In particular, the children of immigrants generally outperform the children of Canadian-born parents in educational attainment by a large margin. Previous studies have attributed this success to the socioeconomic and cultural characteristics of Canada’s immigrants and the efficacy of Canada’s education and social policies.Note 1
While education systems and social policies are broadly similar across regions, the socioeconomic status and ethnocultural composition of immigrants vary considerably. Such regional differences in immigrant characteristics may be associated with regional differences in the educational outcomes of immigrant children. Previous studies have shown large variations in educational outcomes among children of immigrants from different source regions.Note 2 Immigrant groups from various source regions may be attracted to particular Canadian regions because of their preferences for certain regional attributes such as climate, language, pre-existing ethnic communities and labour market niches. The diverse roles of provincial governments in immigration selection and settlement may also lead to differences in the geographic distribution of immigrant groups.
In Quebec, for instance, proportionately more immigrants selected by the province come from Africa, the Middle East, the Caribbean, and South and Central America.Note 3 Since the late 1990s, other provinces and territories have also become actively involved in immigration selection through nominee programs. These programs allow provinces and territories to nominate and select immigrants to meet their needs for population growth and labour supply. So far, provincial nominees have made up a significant share of immigrants to Manitoba, Saskatchewan and the Atlantic provinces.Note 4 From one province to another, nominees vary considerably in socioeconomic characteristics and source regions. Previous studies have shown that provincial nominees in Manitoba have lower education levels than federal skilled-worker immigrants and are far more likely to come from the Philippines, while provincial nominees in British Columbia are disproportionately high-skilled professionals from Western countries.Note 5
This article has two main objectives. First, it examines the regional differences in reading and math scores between immigrants who immigrated to Canada in their childhood (before the age of 15) and the children of two parents who were born in Canada─also called third-or higher- generation children. It does so by combining data from the various cycles of the Organisation for Economic Co-operation and Development’s (OECD) Programme for International Student Assessment (PISA) conducted since 2000 (i.e., 2000 and 2009 for reading scores and 2003 and 2012 for math scores). Data for these years were aggregated to obtain sufficiently large samples of immigrants for each region. Second-generation children, defined as the children of one or two immigrant parents, are not included in this study (see Data sources, methods and definitions).
Secondly, this paper examines regional differences in various educational attainment measures among individuals who immigrated in their childhood (before the age of 15) on the basis of data from the 2011 National Household Survey (NHS). Two particular measures are examined: high-school completion among young adults aged 20 to 24, and university completion among those aged 25 to 29. In both cases, this article further examines whether differences in the educational outcomes of immigrants across regions of Canada are associated with differences in their socioeconomic and ethnocultural characteristics.
The academic skills youth develop in high school are considered their foundation for further development of human capital and full participation in society.Note 6 In a global context, Canadian students perform relatively well in reading and mathematics. Several recent evaluations from PISA have placed Canada among the top participating countries in student performance levels and equity in learning opportunities. Canada, like Australia, stood out among major OECD countries in the population share of immigrant students and math and reading performance in high school. In most OECD countries, immigrant students lagged behind third- or higher-generation students in academic skills, but, in Canada and Australia, this gap was relatively small.Note 7
At the national level, immigrant students in Canada had similar math scores and slightly lower reading scores than third- or higher-generation students. The lower reading levels likely reflect the fact that the mother tongue of many immigrant students is neither English nor French.Note 8
The differences in scores obtained between immigrant and third- or higher-generation students, however, were not the same from one region to another in Canada. Immigrant students in British Columbia and the Atlantic provinces, for example, had significantly higher math scores than third- or higher-generation students (Table 1), while immigrant students had lower math and reading scores than third- or higher-generation students in Quebec, and in Manitoba and Saskatchewan (the latter two provinces have been combined because of small sample sizes).
|Immigrant children||Third- and higher-generation children (ref.)|
|mean||standard error||mean||standard error|
|Math scores, 2003 and 2012 combined|
|Manitoba and Saskatchewan||494Note ***||4.8||514||1.5|
|British Columbia||554Note ***||4.4||526||2.9|
|Reading scores, 2000 and 2009 combined|
|Manitoba and Saskatchewan||477Note **||14.0||522||2.2|
|British Columbia||523Note *||4.4||534||2.8|
Furthermore, interregional differences in math and reading scores were larger among immigrant students than among third- or higher-generation students. Among third- or higher-generation students, math scores ranged from 545 (in Quebec) to 505 (in the Atlantic provinces), a difference of 40 points. Among immigrant students, they ranged from 554 (in British Columbia) to 494 (in Manitoba and Saskatchewan), a difference of 60 points.
Regional differences in math and reading scores among high-school students may originate partially from variations across provinces in school curriculums; the share of public, religious and private schools; teacher training; and school resources.Note 9 Among immigrants, however, interregional differences in characteristics and family background may also play an important role. The extent to which these factors account for regional differences in average test scores between immigrant and third- or higher-generation students will be examined in the next section.
Of all the factors that may be associated with the skill development of high-school students, parental education and student aspirations are especially important.Note 10 In all regions of Canada, parents of immigrant students were more highly educated than parents of third- or higher-generation students. Even so, immigrant parents in Manitoba and Saskatchewan and in Quebec were less likely to have a postsecondary degree than immigrant parents in other regions (Table 2).
|Immigrant children||Third- and higher-generation children (ref.)|
|percent||standard error||percent||standard error|
|Both parents had a postsecondary education|
|Manitoba and Saskatchewan||48.7Note ***||1.3||27.6||1.0|
|British Columbia||62.2Note ***||2.1||30.2||1.9|
|Students expected to finish university|
|Manitoba and Saskatchewan||71.0Note ***||1.2||63.6||1.1|
|British Columbia||84.8Note ***||1.5||58.8||2.1|
Furthermore, in all regions, a higher proportion of immigrant students aspired to finish university than third- or-higher generation students, but again, these proportions varied by region. In British Columbia, Ontario, Alberta and the Atlantic provinces, more than 80% of immigrant students expected to complete a university degree while this was the case for 73% and 71% of immigrant students in Quebec, and in Manitoba and Saskatchewan, respectively.
To evaluate factors contributing to differences in PISA scores between immigrant and third- or higher-generation students, a linear regression model was run. Math and reading scores were included as the dependent variables, while explanatory variables in the model included parental education, student educational aspiration, sex, language spoken at home,Note 11 family structure, parental occupational status, number of books at home, school type, and year. These factors accounted, in varying degrees, for the differences in PISA scores observed between immigrant and third- or higher-generation students in different regions.Note 12
These factors accounted for all of the difference in math scores between immigrant and third- or higher-generation students in Ontario, and for approximately one-half of the difference observed in British Columbia. Differences in the educational aspirations of students, and to a lesser extent, in the educational attainment and occupations of parents, were the main factors accounting for the advantage of immigrant students in British Columbia and Ontario (Chart 1).
Among provinces or regions where immigrant students lagged behind third- or higher-generation students in math scores, speaking neither English nor French at home was a common factor contributing to the lower outcomes of immigrant students. This was the case in Alberta and in Manitoba and Saskatchewan, where most of the gap could be accounted for by background variables. In Quebec, however, only about one-tenth of the 40-point gap between immigrant and third- or higher-generation students was accounted for by background factors.
With regard to PISA reading scores, socioeconomic characteristics accounted for about one-third of the 50-point gap between immigrant and third- or higher-generation students in Quebec, for one-half of the 13-point gap in Ontario, for two-thirds of the 45-point gap in Manitoba and Saskatchewan, and for the entire gap in Alberta (13 points) and British Columbia (11 points). In all regions, the primary factor associated with the explained portion was language spoken at home (Chart 2).
The previous section documented some of the factors contributing to the differences in PISA scores observed between immigrant students and third- or higher-generation students within regions. The focus of this paper now shifts to the differences in PISA scores observed among immigrant students across regions. As shown above, immigrant students in Quebec and in Manitoba and Saskatchewan had lower math and reading scores than immigrant students in other provinces. Such differences may again be related to socioeconomic and ethnocultural characteristics of immigrant students in different provinces.Note 13
These characteristics accounted for about one-third of the lower math scores obtained by immigrant students in Quebec and in Manitoba and Saskatchewan relative to their Ontario counterparts (Table 3), with lower educational aspirations and parental occupational status the most significant factors. Conversely, about one-third of the higher math scores obtained by immigrant students in British Columbia (relative to their Ontario counterparts) were due to background factors, particularly educational aspirations, parental occupation, and parental education.
|Math, 2003 and 2012 combined||Reading, 2000 and 2009 combined|
|observed difference||adjusted difference||observed difference||adjusted difference|
|Province or region|
|Quebec||-23.9Note **||-15.1||-37.4Note ***||-23.7Note **|
|Ontario (ref.)||Note ...: not applicable||Note ...: not applicable||Note ...: not applicable||Note ...: not applicable|
|Manitoba and Saskatchewan||-38.0Note ***||-22.2Note **||-45.6Note **||-29.6Note ***|
|British Columbia||22.0Note **||15.9Note **||0.1||-7.0|
... not applicable
Sources: Organisation for Economic Co-operation and Development, Programme for International Student Assessment (PISA), 2000, 2003, 2009 and 2012.
Differences in reading scores were also partly explained by background characteristics. About one-third of the overall difference between Ontario and Quebec immigrants, for instance, were related to characteristics such as the number of books at home, language spoken at home, educational aspiration, and parental occupation. Background differences also accounted for about one-third of the difference between Ontario and Manitoba/Saskatchewan, with parental occupation and educational aspiration variables the main contributing factors.
Note that the PISA data used in this study does not provide information about the source region of immigrant students. Previous studies show that the educational attainment of immigrant children varies considerably by source region even after adjusting for family socioeconomic status and students’ educational aspiration. The variation by source region likely reflects the possibility that various immigrant groups value education differently and invest varied levels of effort into the education of their children.Note 14 Children of immigrants from East Asia (e.g., China) and South Asia (e.g., India), in particular, tend to have a higher educational attainment than those from Southeast Asia (e.g., Philippines), the Caribbean, Central and South America, and Southern Europe.Note 15
The potential effects of source regions on the regional differences in educational outcomes among childhood immigrants can be examined using the 2011 National Household Survey (NHS), which collected detailed information about the source countries and educational attainment of immigrants.
The NHS does not collect information on math and reading scores. It does, however, collect information about the highest level of educational attainment, allowing for the calculation of high school and university completion rates. In this section, comparisons are again made between third- or higher-generation Canadians and persons who immigrated before the age of 15. High school completion is defined as high school graduation among those aged 20 to 24, while university completion is defined as obtaining at least one university degree among those aged 25 to 29.
In all provinces and regions, and in Canada as a whole, immigrants had higher high-school and university completion rates than third- or higher-generation Canadians. This was also the case in Quebec and in Manitoba and Saskatchewan, even though immigrant students in high school had lower math and reading scores than third- or higher-generation students in those provinces.
Note that some of the 20- to 24-year-olds in the 2011 NHS were in the same birth cohort as those in the 2003 PISA when math scores were evaluated for those at age 15. Similarly, some of the 25- to 29-year-olds in 2011 were in the same birth cohort as those evaluated at age 15 in the 2000 PISA. This suggests that these cohorts of immigrants might have overcome a possible disadvantage in reading skills in high school to eventually become more likely to complete high school and obtain a university degree than their third- and higher-generation counterparts.
Even though immigrants had higher completion rates than their Canadian-born counterparts in all regions, there were significant regional variations in this regard. Young immigrants in British Columbia had the highest high school and university completion rates, while immigrant youth in Quebec and in Manitoba and Saskatchewan had the lowest rates. Specifically, 44% of young immigrants aged 25 to 29 in British Columbia held a university degree, compared with 29% in Manitoba and Saskatchewan and 32% in Quebec (Table 4). The lower rates of university completion observed among immigrants in Quebec, as well as Manitoba and Saskatchewan, mirrored the regional differences in PISA math and reading scores documented above.
|Third- and higher-generation individuals||Immigrants|
|Observed rate||Adjusted rate|
|High school completion among those aged 20 to 24|
|Canada||86.6||93.1||Note ...: not applicable|
|Manitoba and Saskatchewan||81.7||89.0||90.8|
|University completion among those aged 25 to 29|
|Canada||25.7||39.9||Note ...: not applicable|
|Manitoba and Saskatchewan||22.0||29.3||35.3|
... not applicable
Note: Individuals who changed province of residence in the 5 years prior to the survey were excluded from the calculation of high-school and university completion. The adjusted rate is derived from a probit regression model with high-school or university completion as the dependent variable. The explanatory variables include sex, age, the use of an official language at home, and immigrant source regions.
Source: Statistics Canada, National Household Survey, 2011.
As was the case with math and reading scores, some of these regional differences may be attributable to background characteristics. One such factor includes the source region of immigrants, which can be retrieved from NHS data (contrary to PISA, which did not collect that information in every cycle). The NHS, on the other hand, did not collect information on student aspirations or parental education.
East Asia and South Asia were the two leading immigrant source regions for Canada as a whole in 2011, together accounting for 31% of immigrants aged 20 to 29 (Table 5). These source regions were followed by Southeast Asia and West Asia/Middle East, accounting for another 22%. Together the four Asian regions accounted for more than one-half (53%) of all immigrants aged 20 to 29.
|Canada||Atlantic||Quebec||Ontario||Manitoba and Saskatchewan||Alberta||British Columbia|
|Central and South America||8.6||4.4||12.4||8.2||19.3||10.2||4.8|
|Oceania and other||0.6||0.6||0.1||0.3||0.8||1.4||1.7|
|Source: Statistics Canada, National Household Survey, 2011.|
While Ontario was similar to the national average in its composition of immigrant source regions, it had a higher share of young immigrants from South Asia and a lower share of immigrants from East Asia. In British Columbia, East Asians accounted for 41% of immigrants aged 20 to 29, more than twice the national average. In Quebec, the four leading source regions were Africa, West Asia/Middle East, Central and South America and the Caribbean, together accounting for 53% of immigrants aged 20 to 29. In Manitoba and Saskatchewan, Southeast Asians accounted for 22% of immigrants of this age group, more than double the national average; immigrants from Central and South America accounted for another 19% (more than twice the national average).
Such differences in source regions were significantly associated with regional differences in high school and university completion rates. As shown in the third column of Table 4, once regional differences in immigrant source region, as well as age, sex, and use of official language at home were taken into account, regional differences in educational attainment became smaller─especially in the case of university completion.
Accounting for differences in source-region composition, age, sex, and use of official language reduced the overall difference in university completion rates between the province with the highest rate and the province with the lowest rate, from 15 percentage points (between 44% in British Columbia and 29% in Manitoba and Saskatchewan) to 7 percentage points. Differences in immigrant source regions played the largest role in this reduction.
Since the NHS does not contain direct measures of parents’ socioeconomic status and student aspirations, however, it is not clear to what extent the effect of source-region composition reflects differences in family socioeconomic status and individual motivations.
In Canada, immigrant students (who arrived before the age of 15) tend to have similar math skills and lower reading skills than third- or higher-generation students in high school. Immigrants, however, have higher rates of high-school and university completion than their third- and higher-generation counterparts.
The educational outcomes of immigrants were characterized by significant regional differences. For instance, immigrant students in Quebec and in Manitoba and Saskatchewan lagged behind their counterparts in other provinces in their PISA math and reading scores in high school, high-school completion rates at ages 20 to 24, and university completion rates from ages 25 to 29. Some of these differences were related to differences in the background characteristics of immigrants—note, however, that differences in source countries cannot be accounted for with PISA data. Differences in source countries matter—they helped explain a significant part of the regional differences in the university completion rates of young immigrants (according to NHS data).
Regional differences in immigrant source countries may be the result of multiple factors that include preferences for different regional attributes, immigrant selection programs that direct immigrants to specific regions of the country, and interprovincial migration flows among immigrants.Note 16 According to previous research, immigration selection and settlement polices are more likely to affect the destination of immigrants than subsequent interprovincial mobility.Note 17 Differences in source country, along with the effect of background characteristics such as parental aspirations and parental education, will likely continue to influence regional variations in the educational outcomes of young immigrants in the future.
Feng Hou is a senior researcher with the Social Analysis and Modelling Division and Qi Zhang is an economist with the Producer Prices Division at Statistics Canada.
- Date modified: