Each year, millions of children and adolescents fall victim to school bullying across the world (Lian et al. 2021). Bullying seems to be a particularly obstinate problem that is difficult to tackle (Chaux and Castellanos 2015). This is problematic, as being bullied is associated with numerous negative outcomes that persist throughout adulthood, such as depressive thoughts, feelings of anxiety and developing an eating disorder. Overall, victimization is clearly a detrimental experience with harmful consequences that should be avoided as much as possible. School bullying is characterized by a power imbalance between students and, thus, identifying potential sources of power imbalance can be an effective way to prevent school bullying from happening (Chaux and Castellanos 2015). We argue that being retained could be such a predictor, as it has previously been linked with having fewer same-grade friendships (Demanet and Van Houtte 2016) and being stigmatized (Jimerson 2001). Moreover, retainees are often labeled as so-called "slow learners" via a distinct and visible ritual (the process of being retained), which could create the perception that retainees are less intelligent. Therefore, based on the labeling theory (Becker 1963), we expect that being retained can lead to a power imbalance between students (based on stigmatization and social isolation), and thus, increases the likelihood of being bullied. Hence, this study examines the association between grade retention and self-reported victimization. The label remains highly visible during the retention year, but tends to wash away gradually in higher grades (Demanet and Van Houtte 2013). Therefore, it is expected that victimization will be stronger associated with being retained in secondary education than with grade retention in primary education.

Moreover, earlier research has found that being retained becomes a less isolating and less stigmatizing experience in contexts where grade retention is a more common practice (Demanet and Van Houtte 2016). Demanet and Van Houtte (2016), for instance, have found that school retention composition - the percentage of retainees at school - moderates the effect of individual grade retention on the number of same-grade friendships. We expect similar moderation effects of both school- and country retention composition on the association between grade retention and victimization, meaning that retainees would be less likely to be victimized in schools and countries with a high number of retainees.

There are major differences in retention rates between schools and countries (Dupriez et al. 2008; Van Canegem et al. 2021), as key decision-makers hold different views and beliefs on the role and effectivity of grade retention. Some countries, such as the United States, have rigid, standardized educational systems while other countries, such as Belgium, grant teachers and school boards more autonomy regarding the decision whether or not to retain a student (Dupriez et al. 2008). Some countries, such as Spain or Portugal, have high retention rates, which indicates that grade retention is assumed to be an effective educational intervention. Other countries, such as Iceland, barely have students who have to repeat a grade, as those countries do not see merit in the practice. In many countries, grade retention is used as a way to improve the academic homogeneity of class groups (Mons 2007), often in combination with other differentiation methods such as ability grouping and/or tracking. Therefore, each national education system has its own unique form of grade retention, which is why a theoretically substantiated selection of countries is pivotal for a successful cross-national comparison on any potential outcome of grade retention. In this study, we use the heterogeneity management typology of Mons (2007) to conduct a cross-national study on the association between grade retention, school and country retention composition, and victimization in 25 different countries.

Data were obtained from the PISA2018 dataset. Respondents came from a variety of grades, although nearly all of them were 15-year-olds. Hence, findings of the study drew from a same-age comparison, in which it is compared how the victimization for retainees in one grade compared to the victimization for non-retained agemates in higher grades (Ehmke et al. 2010). The final dataset consisted of 159,412 students, across 8,039 schools in 25 countries. The contextualized impact of grade retention upon victimization is assessed by performing cross-national stepwise multilevel analyses. Stepwise multilevel analyses were conducted, using MLwiN. Conventional multiple regression techniques would lead to an overestimation of the effects at higher levels, as educational systems have a nested data structure. This means that students are nested within schools, that are nested within countries. The moderation effect of school and country retention composition was estimated by including cross-level interaction terms. Data were weighed using the W_FSTUWT variable (OECD 2018). This is the final trimmed nonresponse adjusted student weight, which included student-level trimming (OECD 2018). Countries were included in the cross-national analysis if they could be categorized into the heterogeneity management model of Mons (2007). The included countries were Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Luxembourg, the Netherlands, New Zealand, Poland, Portugal, the Slovak Republic, Spain, Sweden, Switzerland, the United Kingdom and the United States.

Throughout all models, there was a significant positive association between being retained both in primary and secondary education, and victimization. Overall, being retained was clearly associated with more victimization which confirmed our hypothesis. As expected, the regression coefficient of being retained in secondary education was substantially larger than the one of being retained in primary education. This indicated that the label of being retained becomes less impactful over the years in terms of being victimized. Initially, we also observed a significant a direct positive association between the country retention rate and victimization, which meant that all students (both retained and non-retained students alike) were more likely to be bullied in countries with high retention rates. This significant association, however, disappeared when heterogeneity management models were integrated. Therefore, the initially observed association was due to national differentiation mechanisms, besides crude retention rates, that were initially unaccounted for. The cross-level term interaction between school retention composition and individual retention in secondary education was not significant; which implies that the association between grade retention and victimization is not impacted by the number of other retainees within school. At the country level, however, a moderation effect did appear, as the cross-level interaction term between country retention composition and individual retention in secondary education was significant. This means that the association between grade retention and victimization becomes less outspoken in countries with a higher retention rate. The findings of this study add to the growing amount of research that questions the desirability and effectivity of maintaining grade retention as a common educational intervention. Moreover, they indicate that the country level and its educational system attributes are more important for understanding the association between grade retention and victimization, despite the closer proximity and visibility of the school context.

Becker, H. 1963. Outsiders. New York, NY: Free Press. Chaux, E., and M. Castellanos. 2015. “Money and Age in Schools: Bullying and Power Imbalances.” Aggressive Behavior 41(3):280-293. Demanet, J., and Van Houtte, M. 2013. “Grade Retention and its Association with School Misconduct in Adolescence: A Multilevel Approach.” School Effectiveness and School Improvement 24(4), 417-434. Demanet, J., and M. Van Houtte. 2016. “Are Flunkers Social Outcasts? A Multilevel Study of Grade Retention Effects on Same-Grade Friendships.” American Educational Research Journal 53(3):745-780. Dupriez, V., Dumay, X., & Vause, A. 2008. “How do school systems manage pupils’ heterogeneity?”. Comparative Education Review 52(2):245-273. Ehmke, T., B. Drechsel, and C.H. Carstensen. 2010. “Effects of Grade Retention on Achievement and Self-Concept in Science and Mathematics.” Studies in Educational Evaluation 36(1-2):27-35. Jimerson, S.R. 2001. Meta-Analysis of Grade Retention Research: Implications for Practice in the 21st Century. School Psychology Review 30(3):420-437. Lian, Q., C. Yu, X. Tu, et al. 2021.”Grade Repetition and Bullying Victimization in Adolescents: A Global Cross-Sectional Study of the Program for International Student Assessment (PISA) data from 2018.” PLoS medicine, 18(11), e1003846. Mons, N. 2007. “Les nouvelles politiques educatives: La France fait-elle les bons choix?” Paris: Presses Universitaires de France. Organisation for Economic Co-operation and Development. 2018. “PISA 2018 results (Volume III): What School Life Means for Students’ Lives.” Paris: OECD Publishing. Van Canegem, T., Van Houtte, M., & Demanet, J. 2021. Grade retention and academic self‐concept: A multilevel analysis of the effects of schools’ retention composition. British Educational Research Journal 47(5):1340-1360.