In the world of political analysis and voting advice applications (VAAs), the claim of “neutrality” is often the first thing you hear. However, academic literature and intellectual tradition teach us that absolute neutrality is a mirage; every choice (from the topics we select to the way we frame a question) carries an inherent perspective. At Palumba.org, we do not claim to be a faceless, neutral entity. Instead, we strive for something more rigorous and valuable: trustworthiness and objectivity through transparency and process.
We uphold the highest internal standards, learning from every iteration of our work and grounding our methodology in the best practices of the sector. Our goal is not to hide our process but to open it up, ensuring that while we may not be in a vacuum, we are fair, balanced, and fiercely objective in our execution. Here is how we build that trust.
1. The Scientific Council: guardians of balance
Every time we deploy our application in a new country (what we call a “mission”) we establish a dedicated Scientific Council. This body acts as our primary safeguard against bias. Composed of trusted partners and local political science experts from think tanks, universities and academia, the Council oversees every step of our content creation process.
Their mandate is clear: guarantee that no ideological bias interferes with our work and ensure the contents of our app are rich. They review our content to ensure that all political topics and positions are portrayed accurately and fairly. By framing debates and issues as starting points for conversation rather than conclusions, the Council ensures that Palumba.org remains a tool for exploration, not manipulation. They have the power to inquire about any aspect of our work, a possibility we also extend to the general public.
2. The process of formulating our statements
The heart of our app lies in the statements users swipe left or right on based on their (dis)agreement. These aren’t just random sentences; they are the result of a constantly improving, iterative process designed to reflect what truly matters to young people.
Crowdsourced & expert-validated
For the 2024 EU Election Mission, we didn’t work in isolation. We received hundreds of contributions from young people and youth organizations across the continent. Simultaneously, our internal diverse team of young policy experts researched the most relevant topics, compiling a massive initial list of over 200 potential statements. As we assess the position of candidates on the subjects, we also follow academic “Voting Advice Application industry standards”. For instance, we worked with the prestigious CIDOB of Barcelona for our 2024 deployment and all team members somewhat involved in editorial processes sign an Ethical Charter on non-partisanship and conflicts of interest.
The selection process
Narrowing this list down is a delicate art. We monitor election campaigns in real-time to identify the most heated and relevant topics. Our approach is grounded in academic guidelines on VAA development, including the work of Veikko Isotalo, who directly advised us during our inception in 2024. Following his guidance, we ensure a mix of topics including:
-
Legal competencies: Issues strictly connected to the powers of the entity being voted on (e.g., the EU Parliament).
-
Emotional & value-based topics: Issues that drive political identity and engagement, even if they are less legislative in nature.
This mix ensures we capture the full spectrum of the political debate. Our Scientific Council reviews the final selection to ensure it is balanced and representative.
3. Transparency in the algorithm: simple and open-source by design
We believe that a “black box” algorithm has no place in sucha core process for democracy. Our match-making system is designed to be as transparent and understandable as possible. We utilize a two-part system:
-
Manhattan distance: A straightforward mathematical calculation of the difference between your answers and a candidate’s positions.
-
Dimensional analysis: We categorize topics into “dimensions” to help users see where they stand on specific thematic areas.
Why this approach? We chose these methods because they are the simplest to explain to the general public while remaining mathematically robust. They work seamlessly with the Likert scale input method we use for your answers. A Likert scale allows you to express not just “yes” or “no,” but the intensity of your agreement or disagreement.
To further guarantee trust, our algorithm is open-source. Anyone with the technical know-how can inspect our code to verify that the match-making is fair and unbiased. You can read more about it here: https://www.palumba.eu/faq
4. Impact assessment: combatting affective polarization and disengagement simultaneously
Our work doesn’t end when the polls close. After the election, we shift our focus to impact assessment. We analyze our data to understand who we reached and, crucially, who we missed. Our goal is to constantly broaden our audience with diversity in mind, ensuring we are not just preaching to the converted.
By mobilizing people around the issues that matter to them, we contribute to a solid, data-driven picture of the election. This empowers further research and serves as our primary weapon against misinformation and polarization. You can read more about this in our Theory of Change article.
5. Our Board of Advisers as an overarching advisory group
While the Scientific Council oversees specific missions, our entire organization is guided by a Board of Advisers. This body is composed of distinguished academic researchers from top universities and prominent figures from civil society (find them here). They do not manage day-to-day operations but help enshrine our values, ensuring that Palumba.org remains true to its mission of empowering youth democracy without compromising on integrity.
A note on Artificial Intelligence
The emergence of AI brings both opportunities and significant challenges for democracy. At Palumba.org, we are not fully neo-luddites; we recognize that AI can aid in internal processes and research efficiency. However, we uphold the same rigorous standards for AI as we do for human-generated content.
We believe AI must be accountable, transparent, and aligned with pioneering global principles on its deployment. To ensure we navigate this new frontier responsibly, we have been in direct contact with Maria Ressa, a Nobel Laureate and UN appointee for the Independent International Scientific Panel on Artificial Intelligence. Their insights help us ensure that if and when we use AI, it serves the democratic process, rather than undermining it.
