On Tuesday morning, the city council sent 18,000 emails inviting people to comment on its new green plan. The message was short, contained a blue button, and promised that every response would count. By six in the evening, the dashboard showed hundreds of submissions and the team celebrated that the consultation was working. Yet nobody at the municipal market had seen the email during their shift; many people at the health centre did not have an address they checked every day; and several members of an association for recently arrived women neither understood the link nor trusted that an answer in another language would be read. The response total did not describe the city. It described those who already had time, language, and confidence to enter through that door.
Follow the route, not just the sent button#
The useful first question is not how many messages went out. It is what a person must do to get from everyday life to a response that has an effect. Consider Nadia, who cleans offices from six to ten and then takes her father to rehabilitation. The notice arrives in an account she opens once a week on her phone. The form asks her to register, confirm an email, read a twenty-seven-page document, and choose priorities by Friday. Nadia knows the square that floods in rain and the unshaded bus stop where she waits, but the design reads her silence as lack of interest.
The useful first question is not how many messages went out. It is what a person must do to get from everyday life to a response that has an effect.
That silence is not neutral data. It is the result of choices about deadlines, channel, wording, identification, and the way information is returned. It also affects Manuel, who uses email but distrusts links after fraud attempts, and Samira, who can explain herself clearly in conversation but does not feel safe writing a technical contribution in English. None of them needs a lesson in citizenship. Each needs a route proportionate to the decision being requested: to know what the plan concerns, to contribute experience without completing an obstacle course, and to receive a sign that their effort did not fall into an automated inbox.
Time is participation infrastructure#
Consultations often measure languages, devices, or accessibility and sometimes overlook the most unevenly distributed resource: unhurried time. A five-day deadline can seem reasonable in an office meeting and impossible for someone who combines shifts, cares for another person, shares a phone, or needs to speak first with neighbours and family. By demanding an individual, immediate opinion, the system favours the response of someone who already has room to read, compare, and write.
Correcting this does not mean holding a consultation open forever. It means publishing an honest timetable. The council can announce stages in advance, offer two or three in-person sessions outside working hours, allow audio or paper submissions, and keep an asynchronous period long enough for groups to talk. Rather than presenting the complete document as the only entry point, it can ask situated questions: where shade is missing, which summer journey feels unsafe, which park is unused and why. Everyday experience does not need to adopt the vocabulary of a planning report to matter.
Language is not a translation tick-box#
Translating the opening message helps, but it does not solve everything contained in a consultation. A machine translation of the email may retain difficult terms, open a form available in only one language, or lead to a document that assumes technical knowledge. More importantly, it may leave untouched the insecurity of someone who fears making a mistake, being misunderstood, or exposing themselves by providing personal data.
A linguistically accessible version explains which decision is open, what has already been decided, and what real room remains. It does not hide the important question behind acronyms and budgets without context. It offers interpretation at sessions, an answered phone line, and the option of an oral response. When contributions need translation, this should happen under visible rules: not to correct a participant's voice, but to bring it into deliberation without losing its meaning. The council can publish how original answers will be preserved, who may access them, and how examples will be anonymised. That clarity protects privacy and trust together.
Trust cannot be manufactured with a slogan saying “your opinion matters.” It grows when the council acknowledges reasonable reasons for distrust: earlier consultations with no feedback, changes that were already decided, forms that seemed to collect data for another purpose, or meetings where only those familiar with the rules spoke. A team seeking to repair that relationship should visit organisations that are not always invited, listen to what failed before, and accept that some people prefer to participate through a trusted group. Representation does not replace every voice, but it can make a bridge when the institutional channel feels distant.
Build several doors to the same place#
The answer is not to abandon automated delivery, but to prevent it from becoming the only door. One participation record can bring together short forms, phone calls, facilitated conversations, paper contributions, and public sessions. For this mix to be fair, every channel must reach the same record and receive the same treatment. If digital comments enter a spreadsheet while notes from a meeting stay in a drawer, plurality is only apparent.
Before beginning, agree on a simple protocol. Each contribution receives a reference, not necessarily a name. Record the channel and, if the person wishes, the neighbourhood or circumstance that helps interpret the issue, without turning that information into an intrusive profile. A team groups responses by theme, separates those requiring clarification, and publishes an understandable summary of what it heard. Observations gathered through mediation must be checkable by those who offered them: an association should be able to confirm that its idea was not reduced to a convenient phrase.
In the green consultation, for instance, it is not enough to count how many people ask for more trees. Record why they ask and what effects they expect: shade on the walk to school, rest for older people, less heat in homes without lifts, play space, access to a drinking fountain. That context prevents a contribution from an area with very little green space from competing in false equality with an aesthetic preference made elsewhere. Deliberation is not adding clicks. It is making the different consequences of a single decision legible.
Count absences without turning them into statistical decoration#
No consultation literally represents everyone, and promising that would be misleading. It can still explain whom it reached, whom it does not know whether it reached, and what it tried to correct. When the process closes, the report should show channels used, times, languages available, districts, and obstacles found. Not to label vulnerable groups from a distance, but to give the next invitation a verifiable basis for improvement.
There is a decisive difference between saying “few young people participated” and asking which part of the process made entry difficult. Perhaps the consultation coincided with exams; perhaps the subject was framed as abstract planning rather than journeys, sport, or heat; perhaps nobody went to the places where young people spend their afternoons. Absence does not prove indifference. It is a hypothesis that needs investigation and feedback. When the team does not know why a voice is missing, it should say so instead of filling the gap with assumptions.
Feedback is when the invitation shows whether it was a relationship or merely a campaign. Before approving the plan, the council can publish which proposals entered, which did not, and why; identify which decisions remain open; and explain the next contact point. It must do this through the same multiple channels used to invite people. A neighbour who left a note at the library should not discover the result only if she searches a municipal website months later. If a proposal does not fit the budget, the response can name a concrete limit rather than use an administrative sentence that closes the conversation.
The practical test is simple: imagine Nadia, Manuel, and Samira want to intervene in the next phase. Will they hear about it in time? Will they understand what is being asked without exposing themselves? Will they find a person and an alternative if the form fails? Will they see what changed because of their contribution? If the answer depends on frequent email, spare hours, and prior trust, participation has not been designed. A selection has been automated.
A fairer green plan begins by admitting that democratic infrastructure includes schedules, translation, mediation, formats, and memory. The next step can be small and testable: before launching the next consultation, bring together five organisations from different neighbourhoods to walk through the invitation, note its barriers, and publish which changes will be made. This does not ask the city to speak with one voice. It asks the institution to stop mistaking the silence created by its own design for consent it never received.
