- Exam-takers with migrant background also receive lower marks
- Even stronger disadvantage at the threshold to top-level scores
- Composition of examination panels an important factor
Berlin, 26 April 2018 - Women score just under 2 percent lower than men in the second state law exam in Germany’s largest state. Among top scores, the gender effect to the detriment of women is particularly pronounced: 12 percent fewer women surpass the key career-relevant hurdle of 9 points. Only those with a score of 9 points or better – a so-called Prädikatsnote, or a score ‘with distinction’ – are admitted to government service, for example. When other factors such as A-level grades, age and examination dates are included in the statistical comparison, the differences are even more pronounced.
Having a migrant background also leads to worse scores. For example, legal trainees who are born abroad and do not have German citizenship score 17 percent lower in the second state exam than German candidates. The probability of achieving a score ‘with distinction’ is as much as 70 percent lower for this group. German-born candidates with a German passport but a "non-German" name are also awarded lower scores on average. The differences also persist when previous grades are taken into consideration in the analysis.
These are the results of a study by Andreas Glöckner (FernUniversität Hagen), Emanuel Towfigh (EBS Universität Law School), and Christian Traxler (Hertie School of Governance) on behalf of the Ministry of Justice of the State of North Rhine-Westphalia (NRW). It is based on a comprehensive data set that includes the exam results of around 20,000 candidates who passed their first and second state exams in NRW between 2006 and 2016. The Ministry of Justice commissioned the study after the authors discovered gender and family background effects in the grading of state law exams in a first study in 2014. The follow-up study underpins and differentiates the results based on a broad set of data.
According to the study, the composition of the three-person panel of examiners has a clear influence on the observed gender effect in oral examinations. Female legal trainees who have the same written grades as their male counterparts have a 2.3 percentage-point lower chance of making the next-higher score on the exam. However, if at least one woman is on the panel, that difference disappears. In mixed committees, men have a marginally worse chance, while women have a marginally better chance, of progressing to the next-higher score. This effect intensifies at the threshold of the grading range ‘with distinction’. While the gender difference for the male-only examination panel is 6 percentage points, the probability that women will receive the next-higher score in mixed commissions increases by 3 percentage points. For men it falls by the same value, completely balancing the results.
From Glöckner, Towfigh and Traxler's point of view, this points to discrimination – even if possibly unconscious – as the cause of the disparities. But examination boards have leverage to counteract the discrimination: 52 percent of exam candidates are women, but 65 percent of examination panels were purely male during the period under review. Only at the end of the period did the proportion of mixed staffing increase significantly. "Female examiners are important for a gender-neutral assessment, and their participation should thus be bolstered," the authors said. Furthermore, they should consider withholding candidates' previous grades from examiners. The significant rise in observed effects around the relevant grading thresholds, in particular the threshold for top distinction, suggests a strategic approach by the examiners: "Aggregating independent written and oral examinations could compensate for errors in the appraisal and thus produce more valid evaluations," according to the authors.
Empirically founded recommendations to counteract effects related to a candidate’s family background, as measured in the study, are more difficult to assess. The number of examiners with a migrant background has so far been so low that statistical statements are not possible. An analogy to the gender effect is, however, possible, which speaks for the use of more examiners with a migrant background. The authors recommend further analysis, especially in this area.
A German summary of the study „Empirische Untersuchung zur Benotung in der staatlichen Pflichtfachprüfung und in der zweiten juristischen Staatsprüfung in Nordrhein-Westfalen von 2006 bis 2016“ by Andreas Glöckner, Emanuel Towfigh and Christian Traxler can be found here. The full study (62 pages) can be downloaded here.
The Hertie School of Governance is a private university based in Berlin, Germany, accredited by the state and the German Science Council. It prepares exceptional students for leadership positions in government, business, and civil society. Interdisciplinary and practice-oriented teaching, first-class research and an extensive international network set the Hertie School apart and position it as an ambassador of good governance, characterised by public debate and engagement. The school was founded at the end of 2003 as a project of the Hertie Foundation, which remains its major partner. www.hertie-school.org