The numbers speak an unmistakable language. The success rate of JRC ASTs in the certification procedure is considerably below that of ASTs from other DGs. While JRC’s ASTs accounted for approximately 9.5% of all ASTs in the Commission in 2019 , henceforth considered to be the benchmark, their share of certifications is much lower and persistently so (Figure 1): Except for a one-off-high in 2016 with a share of 7% (although still below 9.5%), the share of certifications in other years was much lower. These facts point to a significant people and talent management problem in the JRC, all the more so as the remoteness of most JRC sites significantly reduces mobility and thus career opportunities compared to staff in Brussels. So “escaping” from a job or a DG where certification, or more generally career progress, is difficult, say because there are not enough AD posts, is even more challenging than elsewhere.
What are the possible reasons for the alleged “under-performance” of JRC? Figure 2 offers some clues.  To begin with, it is quite clear what can be excluded as possible explanations: There does not seem to be a lack of eligible candidates from JRC, as the JRC share at that stage is still very close to the benchmark. (ii) Nor are not enough JRC ASTs admitted to the interview. While only between 20% and 34% of candidates are selected to participate in the interview in the first place, the JRC share among those remains roughly the same compared to the JRC share in all eligible candidates (9.3% vs. 8.4% on average). The big drop (or should we say carnage?) occurs thereafter. The share of JRC ASTs successfully passing the interview and being admitted to the training drops to a meagre 3.7% on average. Since thereafter JRC candidates perform, if anything, better, the problem lies obviously with the panel interview. Also the most recent figures for 2020 confirm this general picture, albeit with some nuances. There were 351 candidates, out of which 37 came from the JRC (10.5%). Selected for the interview were then 102 colleagues of which 5 were from JRC (4.9%), but only 3 of these (6% of the total) successfully passed the interview phase. Unlike in previous years, the JRC appears to have been rather strict in selecting candidates for interviews, thus anticipating the outcome of the interviews.
So if the interview is the highest hurdle, why is this so? According to the available evidence, the questions posed and the expected answers in that interview are clearly biased in favour of people with knowledge of, and experience in, working close to the policy cycle. Regardless of their brilliance as (mostly) scientists, the JRC’s ASTs therefore face a structural problem when it comes to answering questions from a rather different area and explaining the added value they bring as AD officials to the Commission as a whole (which, one should not forget, is mainly in the business of policy making).
Of course, the talent management team of the JRC provides support and help, for instance by offering mock interviews for candidates. However, this does not change the fact that JRC ASTs have to go through an exercise which is not only unrelated to much of their previous work, but also to their likely future work as an AD official in the JRC. Teaching candidates the “correct” answers may help to increase the JRC share, but at the cost of being not altogether open about what they expect to do in the JRC.
The certification exercise is therefore in stark contrast to other types of AD recruitment, most notably, open external competitions. These are very much tailored to the expected needs of the service in terms of experience and qualifications. For instance, candidates in the recently finished AD competitions for Scientific Research Administrators could choose between no less than six different fields. And while the notice of competition  made it clear that some knowledge of the policy cycle would be an advantage, the specific competencies required came largely from the scientific field. Against this background, it is difficult to see why a one-size-fits-all approach to certification is appropriate, especially since open competitions for ASTs also cover a wide range of competences and qualifications.
When confronted with this problem by Generation 2004 staff representatives and others, JRC management has shown a certain openness to these arguments, but then put forward the resistance of DG HR to any change in certification as an explanation why nothing can and will be changed. One can therefore only hope that Mrs Souka’s well-deserved retirement and the arrival of the new Director General of JRC will give new impetus to the matter.
 Sources: https://ec.europa.eu/info/sites/info/files/european-commission-hr-key-figures_2019_en.pdf , https://myintracomm.ec.europa.eu/sg/Documents/governance_departments_guide.pdf . For previous years, the share of JRC ASTs in all Commission ASTs is not readily available.
 These figures have been obtained from administrative notices between 2014 and 2019 where lists of successful candidates for each stage together with their home DG have been published.