To the attention of the Director of EPSO and Members of the Management Board

András Baneth's picture
András Baneth

To the attention of the Director of EPSO and Members of the Management Board

Subject: Towards a stronger, fairer, and future-proof EU selection system

Dear Director,

Dear Members of the Board,

First, allow us to express our appreciation for EPSO’s mission and the difficult balance it must strike between fairness, transparency, political expectations, and operational realities. The AD5’s relaunch and the onboarding of the new test provider is a welcome sign that EPSO is overcoming its recent challenges, and it reflects your effort to meet changing institutional needs and candidate expectations.

At the same time, as an entity working at the intersection of EU careers and occupational psychology, we would like to share a few reflections and proposals to help future-proof the system and increase both its credibility and effectiveness in selecting Europe’s most capable civil servants.

1. Test what matters, and rank accordingly

  • The current approach to weigh the EU knowledge or field-related tests as the primary ranking mechanism risks distorting selection outcomes. EU knowledge is undoubtedly valuable as it demonstrates candidates’ genuine interest and baseline understanding of how the EU functions. From a selection science perspective, however, its usefulness is largely limited to reflecting candidate motivation. While motivation is certainly important, it is not the strongest predictor of future job performance. Making EU knowledge the primary ranking criterion risks over-emphasising motivation at the expense of other critical competencies such as problem-solving, adaptability, teamwork, and communication. Identifying the most suitable candidates is inherently more complex and requires a multidimensional assessment approach that ensures both fairness and scientific validity. Using the terminology of occupational psychology, the use of EU knowledge tests in this way produces invalid results because such tests have very limited predictive validity (they don’t really predict future job performance of candidates). 
  • Cognitive tests (abstract, numerical and verbal reasoning) are consistently shown to be among the strongest predictors of future job performance, capturing candidates’ capacity to analyse, reason, and learn quickly. If the system intends to focus on presenting the most suitable candidates who are most likely to perform well in their future job roles, ranking should be done on the basis of cognitive tests. 
  • The new EPSO system is notably missing any behavioural or competency-based assessments. This is a significant concern from the effectiveness of the selection system, as these methods are essential for evaluating how candidates are likely to perform in real job contexts. While cognitive ability tests capture problem-solving and reasoning skills, behavioural assessments such as situational judgement tests (SJTs), e-tray exercises, and structured interviews go further by examining interpersonal skills, adaptability, teamwork, and decision-making under pressure. These dimensions are critical for the roles of EU officials, which demand not only analytical ability but also collaboration, leadership, and resilience in complex environments. EPSO itself has a strong history of using such tools. E-tray exercises and SJTs were long-standing parts of the assessment centre stage, as well as structured interviews were a core component in evaluating candidates’ competencies against the relevant competency framework. The removal of these tools risks narrowing the assessment focus and undermining both the predictive validity (i.e. “is EPSO really choosing the best candidates?”) and the fairness of the selection process. A return to the use of competency-based methods would strengthen the credibility of the system and ensure a more holistic evaluation of candidates’ suitability. We see a clear value in re-exploring validated situational judgment tests or short personality inventories: not for exclusionary (selecting-out) purposes, but to enhance institutional fit and improve diversity of work styles.
  • At present, there is little clarity on how EPSO tests are designed, validated, and reviewed for bias. It is also unclear whether they are supported by published evidence showing that different versions are of equal difficulty. Without this, test validity is weakened and the risk of socio-demographic bias increases. Greater transparency and systematic checks are necessary to ensure that no groups, wherever they may be living in the European Union, are excluded or discouraged from taking part in the selection procedure. Small procedural changes (e.g. test instructions, interface design, retesting opportunities) are often sufficient to mitigate these risks.

2. Improve transparency and predictability

  • Candidates would welcome more clarity from the outset on why any given test is used and why any given competency is assessed. This improves motivation, reduces test anxiety, and lowers “self-deselection” among talented candidates.
  • More information on the job preview and clarity on the post-competition roadmap would also improve alignment between candidate expectations (including timelines) and institutional realities. This would strengthen candidate engagement and confidence.

3. Consider modernising assessment modalities

  • Technology allows for assessment formats that are both valid and practical.
  • Virtual inbox (e-tray) exercises - once a core element of EPSO competitions - and structured video-based simulations, when reviewed by trained assessors, can effectively capture key behavioural competencies and provide a strong complement to cognitive testing.
  • The careful and ethical use of AI can support (but not replace) human judgment, particularly for test item consistency checks, cheating prevention, and, as EPSO’s roadmap had already mentioned, test generation.
  • Importantly, human review needs to remain central in any high-stakes selection process.
  • Complementary to this, factoring in how to minimise cheating using AI (using methods including instructional format, content delivery methods and technology monitoring tools) will remain paramount going forward.

We strongly believe that EPSO’s work is vital not just for the EU institutions, but for the broader European project. A selection system that is psychometrically robust, socially fair, and aligned with institutional needs can help ensure that talented candidates from all Member States, especially underrepresented ones, see EPSO not as a barrier, but as a trusted and fair gateway.

I hope these ideas contribute to your ongoing work and that our comments are received as constructive towards your overall goal of improved selection processes for EU competitions.

We at EU Training and our team of psychometric experts are available to provide more detailed input in any way that supports your efforts.

With respect and commitment to a stronger EU public service, best regards,

András Baneth & László Zlatarov

EU Training

Ben Williams CPsychol, MA, MSc 

CEO of Sten10 Ltd., Former Chair of the Association for Business Psychology

EPSO news