A public authority can give a lengthy response without answering the question it was asked. If someone challenges the justification for a criterion used in their case and receives an explanation of the software’s accuracy, they will have obtained information. Their objection remains untouched. Democratic debate about algorithms begins with recognising that difference: less eye-catching than publishing code, and rather more important to someone who needs an answer.
Rodrigo Castillo Murillo and Francisco Figueroa Ogalde addressed the issue in “La democracia frente al algoritmo”, published in El Dínamo on 22 June 2026. The linked Spanish column offers access to their argument. Source: El Dínamo
Knowing how it works does not settle its purpose
Consider a hypothetical municipality allocating time at a sports facility according to predicted attendance. The software accurately identifies which activities will attract larger crowds. A small community group repeatedly comes last. It can accept that the estimate is correct while questioning whether maximising attendance should determine the entire allocation. The municipality might also need to consider access for groups with fewer alternatives. Better model training will not make that dispute disappear.
The prediction helps establish a consequence. The allocation rule determines how much that consequence matters compared with others. If both operations are presented as a single technical answer, a political choice acquires the appearance of an inevitable result. The official may act in good faith and believe an objective criterion is being applied. The issue is who decided to make it decisive and how that choice can be examined.
A complaint must be able to change something
Publishing rules makes them knowable. Challenging them also requires an institutional route. The group needs to identify an authority able to consider its objection, the relevant evidence and whether a reply will arrive before the season ends. A form that collects every opinion without giving anyone the power to reconsider the allocation can produce an impressively orderly archive of powerless disagreements.
Not every objection will be justified. The municipality might defend attendance as the criterion for a programme specifically intended to reach large numbers, or reserve other time slots under different rules. Democratic decision-making does not mean satisfying every preference. It requires sustaining and reconsidering the reasons for exercising power over others. The explanation must address the claim presented, rather than merely certify that the program did exactly what it was asked to do.
The advantage missing from the indicator
Automation can improve this process. It can identify unused slots, compare options and reduce the administrative work required to investigate exceptions. That gain creates an opportunity to respond better. If it is used only to process more applications, it can also widen the distance between a decision and those trying to understand it. The same technical saving allows very different institutional arrangements.
The argument deserves attention because it avoids a false choice between accepting the algorithm and returning to an idealised bureaucracy. A human decision can also be arbitrary, opaque or difficult to challenge. The relevant comparison concerns what each process provides for understanding the criterion, disputing its application and obtaining a justified correction. If modernisation preserves the service window while removing anyone behind it who can resolve a case, the procedure will be new; powerlessness will have a better interface.