In Portugal, the annual official data on pupils of immigrant origin, published by the Directorate-General for Education and Science Statistics of the Ministry of Education and Science (DGEEC/MEC), specifically refers to citizenship status (Cândido and Seabra 2019; Seabra and Cândido 2020). However, this data overlooks pupils of immigrant origin who have Portuguese citizenship or those who have acquired the host country’s citizenship (children of immigrants). Given Portugal’s favourable nationality policies promoting immigrant naturalization, it is apparent that the official data underestimates pupils with an immigrant background.

Grouping a large proportion of pupils with and without an immigrant background under the category ‘nationals’ hinders the identification of long-term educational integration and the comparison of inequalities in educational opportunities for immigrants and their children, relative to their native peers. This approach fails to provide insights into the effectiveness of school systems in supporting the growth and development of immigrant pupils (OECD 2019). Consequently, a comprehensive understanding of the reality of migration in Portuguese schools becomes challenging, potentially leading to a positively or negatively biased portrayal and contributing to the formulation of inefficient public policies.

Furthermore, official statistics have scarce information on academic performance, only displaying grade transition rates by national origin (i.e. citizenship). These are limited indicators of academic performance because they only give a snapshot of performance at a particular moment in time.

In this article, we intend to overcome the abovementioned problems using the recently available student-level microdata provided by DGEEC/MEC. This microdata has variables containing information on the student and parents’ countries of birth, enabling us to accurately identify pupils with an immigrant background, as well as to look at students’ school path by creating a proxy variable on the number of times each student has repeated a grade in the past. This constitutes an approximation to a longitudinal analysis of inequalities concerning the degree of system-level social and ethnic selectivity.

Our goals are as follows:

1. Given the paucity of studies of national scope encompassing pupils with an immigrant background, this article’s analysis is guided by two primary objectives.  First, we aim to go beyond the limitations of citizenship classification and understand the advantages of employing a more comprehensive classification scheme. To achieve this, we compare enrolment and academic performance using categories based on citizenship (national/foreign) and those based on their parents’ country of birth (immigrant origin/native).
2. Acknowledging that research in Portugal is limited to and primarily concentrated with children of immigrants as a whole (Mateus 2022), our objective is to examine migrant heterogeneity by analytically deconstructing the assumed homogeneity within this category. In doing so, we introduce subcategories related to generational status (first and second generations), type of ancestry (parentage of mixed origin versus single origin) and national origins.
3. Considering that the national origins groups are categorised according to student citizenship in the official statistics, we aim to compare how enrolment and academic achievement between and within national origins vary according to the criteria chosen to delimit the group (student citizenship or parents’ countries of birth). We also aim to compare the academic achievement of national origins by generational status and type of ancestry.

Lastly, since socioeconomic contexts also explain differences in academic performance, we use “student’s socio-economic index” to explore whether the gaps we identified in these comparisons persist.

The data used was provided by DGEEC/MEC, within the scope of the ‘(In)Equalities in the school paths of descendants of immigrants’ project that is currently underway at CIES-Iscte. The data cover pupils enrolled in primary and secondary education in state schools in continental Portugal during 2021-22 academic year. This data allows us to adopt an extensive methodological approach never before used in Portugal to study the educational paths and achievements of pupils with an immigrant background. Our analysis focuses on pupils in Portugal enrolled in the 10th grade in state schools of continental Portugal. Our aim is to analyse how using different social categories related to migrant status affects the conclusions drawn about gaps in academic performance. The exercise involves comparing indicators based on categories determined by citizenship (foreign/national) and categories based on immigration status (immigrant background/native). In addition to exploring the effects of these two different ways of categorizing pupils, we conducted a more detailed analysis of pupils of immigrant origin through new categorical distinctions, namely generational status (first-generation/second-generation), type of ancestry (single origin/mixed origin – within the latter we highlight those with parentage of mixed origin with one of the parents born in Portugal), and national origins (those with at least 100 students enrolled in the 10th grade). This analysis enables us to examine migrant heterogeneity and uncover differences in academic performance among pupils with an immigrant background not yet known in Portugal. To analyze the intersection of social and national inequalities, we compare students with different migration status and national origins controlling for ‘student’s socioeconomic index’. This index is created by a Multiple Correspondence Analysis (MCA) with three input variables: parents’ educational level, social class (occupation and employment status), and economic support (ES). The former involves attributing the highest educational level available between mother and father to the family unit. The latter involves a combination of both parents’ employment status and job occupation, to derive a family-level categorization of social class that distinguishes students according to their family’s proximity to culturally and economically valued economic spheres (Mauritti et al., 2016), which give them an educational advantage. ‘Academic performance’ is measured by the number of retentions during pupils’ academic path. This indicator is a proxy variable, calculated by determining the difference between a pupil’s age and modal age in each schooling level (or expected age of attendance).

Preliminary findings reveal previously unknown differences in academic performance among pupils with an immigrant background in Portugal, challenging traditional understandings. It is clear that: considering only the citizenship of the pupil underestimates the representation of pupils of immigrant origin; second-generation pupils often exhibit comparable or superior academic paths; and pupils with parentage of mixed origin, especially those with a native parent, demonstrate a significant advantage in academic performance. These differences tend to persist when accounting for social conditions. Analyzing academic achievement by national origin reveals heterogeneity that is hidden in the broadest categories. We identified four homogeneous subsets using the distribution of no. of retentions by national origins: (i) one characterized by a low number of failures during the school path by the time they reach the 10th grade ; (ii) a second one also composed of national origins where most students do not have any failures, but this share is lower than in the first subset; (iii) a third subset characterized by high levels of school failure, where 50% of students achieve 10th grade with at least 1 retention; (iv) and a fourth one marked by aggravated failure (two or more retentions). Although the reasons for these gaps remain unknown, preliminary evidence suggests that the answer may lie in the combination of national origins with parental education, generational status, and type of ancestry. However, three national origins, namely Santomean, Cape Verdean, and Guinean, deviate from this trend. They have intermediate proportions of second-generation students with mixed-origin parentage and Portuguese citizenship but exhibit poorer academic outcomes compared to other origins with similar characteristics. These national origins share a common aspect in that they are formerly-colonized countries by Portugal. A better understanding of this reality may help reduce existing stigmas and clarify the existence of processes of institutional racism in Portuguese schools.

Mauritti, R., Martins, S. da C., Nunes, N., Romão, A. L., & Costa, A. F. da. (2016). The social structure of european inequality: a multidimensional perspective. Sociologia, Problemas e Práticas, 81, 75–93. https://journals.openedition.org/spp/2339 Cândido, A.F. and Seabra T. (2019), ‘Os alunos de nacionalidade estrangeira no sistema educativo português: matrículas e modalidades de ensino’/ ‘Pupils with foreign nationality in the Portuguese education system: enrollment and type of curriculum track’, Observatório das Desigualdades - Estudos, ISCTE-IUL, CIES-IUL. Seabra, T. and Cândido, A.F. (2020), ‘Os alunos de nacionalidade estrangeira no ensino básico e secundário em Portugal Continental (2011/12 a 2016/17): taxas de aprovação’/ ‘Pupils with foreign nationality in basic and secondary education in mainland Portugal (2011/12 to 2016/17): approval rates’, Observatório das Desigualdades - Estudos, ISCTE-IUL, CIES-IUL. Mateus, S. (2022), ‘Blending ahead: The advantages of young people of mixed origin in Portuguese compulsory schooling’, Globalisation, Societies and Education 20 (5): 571–89. OECD (Organization for Economic Cooperation and Development) (2019), PISA 2018 Results, Vol. II: Where All Students Can Succeed, Paris: OECD.