Nationally and internationally, grade retention is a highly controversial measure to homogenise students with different competence levels. In Germany, some federal states (e.g., Hamburg and Berlin) have already abolished grade retention. In Bavaria and Bremen, however, grade retention rates are above the national average (2.3%, see Statistisches Bundesamt, 2018). In an international comparison, the rate of German pupils who have been retained at least once in the course of their educational career is above the OECD-average (Germany: 19.6%, OECD: 12.2 %; ꭓ² = 29558.56, df=1, p<.001; own calculations). Empirical evidence on the effectiveness of grade retention is still insufficient. International studies (Goos et al., 2021) showed that there are short-term improvements in performance after being retained, but they decrease in the medium and long term. Especially in highly hierarchically structured education systems such as Germany, Belgium, the Netherlands or Switzerland, grade retention is least effective (ibid.). In Germany, only a few reliable studies allow concrete statements about the effectiveness of grade retention: Positive effects of grade retention on students’ performance development were not proven (Ehmke et al., 2017; Fabian, 2020). Beyond this, Fabian (2020) also showed that there was no significant improvement in school grades of retained students. Marsh et al. (2017), however, found an increase in retained students’ math performance after repetition. With regard to the achieved school-leaving qualification, Bellenberg (1999) showed that grade retention is very often associated with school dropout and/or downward change of school track, thus reducing the probability of achieving higher school-leaving qualifications for repeating students. Demski and Liegmann (2014) reported only minor differences between repeaters and promoted students. Given the theoretical assumptions of credentialism (Bills, 2003), DiPrete et al.'s (2017) findings from an international comparison are noteworthy: In Germany, school-leaving certificates are particularly important for future success on the training and labour market. At the same time, objective competencies and obtained certificates are often incongruent (Brändle & Pohlmann, 2021). Empirical findings showed that particularly students with low qualifications successfully enter the training market if they have good grades and high educational aspirations – regardless of their cognitive and social skills (Holtmann et al., 2017). Due to the strong correlation between school-leaving qualifications and success on the training and labour market, the present study investigates whether grade retention has advantages or disadvantages for retained students. Previous research findings suggest that being retained might lead to lower school-leaving qualifications.

Data basis for the present analyses was the German National Educational Panel Study (NEPS, starting cohort 3; Blossfeld et al., 2011). The sample was initially representative of Grade 5 students in Germany and consisted of n = 6,491 students. For the present analyses, we excluded students in school tracks where different school-leaving certificates can be obtained (i.e., students from tracks with several educational programmes, comprehensive tracks). Students for whom no information on the attended school track was available were also excluded from analysis. This resulted in an analysis sample of n = 4,371 students, 118 of whom were retained in Grade 7 (2.5%). We treated missing values using multiple imputation (m = 55 ) in R 4.2.2 (R Core Team, 2023) via the package mice (van Buuren & Groothuis-Oudshoorn, 2011), accounting for the clustered data structure and the frequently non-normal data distribution. Based on this analysis sample, we calculated propensity scores using the Rubin Causal Model (Rubin, 1974) and conducted propensity score matching (Rosenbaum & Rubin, 1983) based on objective competencies in reading and math before grade retention as well as information based on key background characteristics provided by students and parents. Grade repeaters were then matched with non-repeaters (full matching, caliper = .10). This allowed us to compare retained students with similar students regarding key background characteristics, but who were regularly promoted. The highest achieved school-leaving certificate, operationalised by CASMIN (König et al, 1988), was then analysed visually. In addition, we created dichotomous dummy variables for (a) a qualification lower than the usual qualification in the respective school type, (b) a qualification appropriate to the school type (i.e., Certificate of Secondary Education at lower tracks [Hauptschule], General Certificate of Secondary Education [GCSE] at intermediate tracks [Realschule], A-level [Abitur] at academic tracks [Gymnasium]) and (c) a qualification higher than the usual qualification at the respective track (i.e., GCSE at lower tracks or A-level after being retained in intermediate tracks). We conducted logistic regressions to analyse the effect of grade repetition on the adequacy of the school-leaving certificate. Results show that a grade retention seems to reduce the chance of achieving a track-equivalent qualification (OR_fit = .92, p = .089). Grade retention had no influence on the chance of achieving a higher qualification than usual in the respective school track. However, grade repetition increased the chance of obtaining a lower qualification (OR_lower = 1.12, p < .001).

In summary, we found that students who were retained in Grade 7 were less likely to achieve a school-leaving certificate that is appropriate to their attended school track. Also – in line with findings from Bellenberg (1999) – the risk of achieving a lower school-leaving certificate increased when students were retained. Since there is some evidence that grade retentions do not lead to better grades (Fabian, 2020) – which is one of the most important goals of grade retention – the findings of Holtmann et al. (2017) become even more important: Even with low qualifications, but good grades and high educational aspirations, young adolescents’ successful transition to the training and labour market is more likely. In that regard, as can be assumed based on the findings of the present study, grade retention fails its goal to help students strengthen their academic outcomes and their chances of obtaining a track-adequate school-leaving certificate. Thus, further investigation is needed to analyse whether students with lower school-leaving qualifications than appropriate to the respective attended school track reach this lower qualification at least with better grades. The reason why previous studies (e.g., Demski & Liegmann, 2014) did not find differences in educational attainment for repeaters and non-repeaters might be that in the past decades, “irregular” – i.e., non-linear – trajectories became more common. Thus, the achievement gap between repeaters and non-repeaters after compulsory education might be narrowed by further training. In their study, Demski and Liegmann (2014) used retrospective information of participants, so the effect of further training could be accounted for. However, we could not address these effects with the data of the present study, yet it will hopefully be possible with the ongoing studies of the NEPS in the future.

Bellenberg, G. (1999). Individuelle Schullaufbahnen: eine empirische Untersuchung über Bildungsverläufe von der Einschulung bis zum Abschluss. Weinheim: Juventa. Bills, D. B. (2003). Credentials, signals, and screens: Explaining the relationship between schooling and job assignment. Review of Educational Research, 73(4), 441-469. Blossfeld, H. P., & Von Maurice, J. (2019). Education as a lifelong process (pp. 17-33). Wiesbaden: Springer Fachmedien. Brändle, T., & Pohlmann, B. (2021). Alles nur eine Frage der Kompetenz? Leistungs- und Chancengerechtigkeit bei der Vergabe von Schulabschlüssen und Abschlussnoten. Zeitschrift für Soziologie, 50(1), 58-77. Demski, D., & Liegmann, A. B. (2014). Klassenwiederholungen im Kontext von Schul- und Berufsbiographien. In: A. B. Liegmann, I. Mammes & K. Racherbäumer (eds.). (2014). Facetten von Übergängen im Bildungssystem. Nationale und internationale Ergebnisse empirischer Forschung (pp. 173-189). Münster: Waxmann (2014) DiPrete, T. A., Eller, C. C., Bol, T., & Van de Werfhorst, H. G. (2017). School-to-work linkages in the United States, Germany, and France. American Journal of Sociology, 122(6), 1869-1938. Ehmke, T., Sälzer, C., Pietsch, M., Drechsel, B., & Müller, K. (2017). Kompetenzentwicklung im Schuljahr nach PISA 2012: Effekte von Klassenwiederholungen. Zeitschrift für Erziehungswissenschaft, 2(20), 99-124. Fabian, P. (2020). Leistungskonsolidierung, Leistungssteigerung-oder etwas ganz anderes? Die Effekte einer Klassenwiederholung auf die Leistungsentwicklung. Münster: Waxmann. Goos, M., Pipa, J., & Peixoto, F. (2021). Effectiveness of grade retention: A systematic review and meta-analysis. Educational Research Review, 34, 100401. Holtmann, A. C., Menze, L., & Solga, H. (2017). Persistent disadvantages or new opportunities? The role of agency and structural constraints for low-achieving adolescents’ school-to-work transitions. Journal of Youth and Adolescence, 46, 2091-2113. König, W., Lüttinger, P., & Müller, W. (1988). A comparative analysis of the development and structure of educational systems: Methodological foundations and the construction of a comparative educational scale. Mannheim: Universität Mannheim, Institut für Sozialwissenschaften. Marsh, H. W., Pekrun, R., Parker, P. D., Murayama, K., Guo, J., Dicke, T., & Lichtenfeld, S. (2017). Long-term positive effects of repeating a year in school: Six-year longitudinal study of self-beliefs, anxiety, social relations, school grades, and test scores. Journal of Educational Psychology, 109(3), 425-438. R Core Team (2023). R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing. Rosenbaum, P. R., & Rubin, D. B. (1983). The central role of the propensity score in observational studies for causal effects. Biometrika, 70(1), 41-55. Rubin, D. B. (1974). Estimating causal effects of treatments in randomized and nonrandomized studies. Journal of Educational Psychology, 66(5), 688-701.