A provisional solution changes its nature when the calendar stops questioning it. What began as a response to an emergency can become, without an explicit decision, the ordinary way of allocating space, data, money, or attention. A community that tests the future does not confuse continuity with success: it sets a review date from the start, preserves alternatives, and requires fresh reasons before renewing something that was supposed to last only briefly.
Temporary measures are valuable because they allow action under incomplete information and learning at a human scale. Trouble begins when the initial urgency erases the conditions for exit. A pilot gathers contracts, routines, expectations, and dependencies; each additional month then looks less like a choice and more like an unavoidable consequence. The professional question is not whether every solution must end, but what evidence distinguishes a response that remains necessary from infrastructure maintained through inertia. That distinction protects both innovation and public trust: an initiative able to explain its duration shows that it still serves a shared purpose.
Duration is also a design decision#
A starting date is usually recorded, while the closing date is expressed as a vague intention. That asymmetry favours permanence. When nobody knows who can close a measure, what evidence that person must present, or what will happen to people who depend on it, passing time decides in place of the community. Temporariness needs an architecture: an identified owner, renewal conditions, a decommissioning budget, rules for accumulated data, and an operational alternative. Without those pieces, a sunset clause is merely a sentence at the end of a document.
A starting date is usually recorded, while the closing date is expressed as a vague intention.
Evaluation must compare the measure with the current problem, not only with the situation on the day it was approved. The emergency, affected population, cost, supplier, or capacity of ordinary services may have changed. Harms that were initially invisible may also have appeared. A serious review asks which benefit continues, how it is distributed, which burdens have become normal, and what less intrusive option is now available. Time then stops functioning as an argument —“it has always been done this way”— and becomes evidence that requires another decision.
Case 1: a temporary square that reorganises a neighbourhood#
A municipality temporarily closed a street to traffic to enlarge a public square during a health emergency. The intervention created room for play, shade, and local commerce, but it also displaced delivery routes and made access harder for some people with reduced mobility. Two years later, the barriers remained because the space was popular and removing it looked like a reversal. The mistake was not keeping it; the mistake was doing so without assessing new consequences. The municipal team documented periods of use, accessibility, noise, deliveries, and residents' views. It then tested a loading point, level paving, and access hours. Renewal was approved for a defined period, with a budget for corrections and another public review. The solution ceased to be a tolerated exception and became an explicit urban decision.
The case shows that expiry does not always mean dismantling. It means recovering the ability to choose. The community could preserve benefits, recognise those bearing concentrated costs, and modify the design without defending it as an indivisible whole. The review date gave a voice to experiences absent during the emergency and prevented aggregate popularity from hiding specific barriers.
Case 2: a school platform introduced for a few weeks#
A school activated a commercial platform to sustain remote lessons. The initial authorisation covered a few weeks, but accounts, alerts, and activity records remained in use years later. Teachers and families now depended on the channel, and changing it seemed too costly. An internal audit found that nobody could state precisely what data was retained, which communications had to pass through the platform, or how a family without a stable connection could participate. The school leadership did not order immediate withdrawal. It first separated essential functions from convenient ones, requested deletion schedules from the supplier, opened an equivalent telephone channel, and piloted an institutional tool with two year groups.
At the end of the term, the school council compared teacher workload, access, incidents, data protection, and communication quality. It decided to close some modules, migrate others, and retain one specific function for six months. Every continuation received an owner and a date. The transition demonstrated that past investment is not sufficient evidence for future renewal and that careful decommissioning is part of design.
Case 3: an emergency neighbourhood network that gains lasting power#
After severe flooding, a neighbourhood network created a list of older residents, available vehicles, and medical needs to coordinate help. The register was crucial during the first days. Months later, some associations wanted to retain it for future emergencies, although its data was outdated and access depended on two volunteers. Good intentions did not remove the risk: a tool of solidarity was becoming sensitive infrastructure without common rules. The network brought together social services, civil protection, neighbourhood groups, and people included in the register. They agreed to delete clinical details, verify consent, restrict access, and transfer ongoing coordination to an accountable public body.
The community kept an activation protocol and regular exercises, not an indefinitely open database. It also required every emergency to have a closing date, an incident report, and a contact for corrections. The learning survived, but the exception did not. That outcome protected the dignity of people who had requested help and prevented the availability of information from replacing the need to justify its use.
Three tests before renewal#
The first test concerns necessity. The accountable owner must describe the present problem using recent evidence and explain why ordinary services are still insufficient. The second concerns proportionality: it compares benefits with costs, unequal effects, loss of privacy, displaced work, and less demanding alternatives. The third concerns reversibility: it checks whether the community can stop the measure without losing rights, documents, support, or essential knowledge. An initiative that works only if it never ends is not genuinely provisional.
These tests need numbers and narratives. Averages reveal scale, but an interview may show why a person drops out, which invisible task falls on a professional, or how a nominal alternative fails. Absences must also be recorded: who does not participate, who cannot use the channel, and who stops complaining. Evidence is not meant to manufacture unanimity. It is meant to make disagreement visible and require the authority to explain why it accepts some burdens, corrects others, or decides to end the measure.
Expiry, review, and renewal are not the same#
Expiry interrupts authorisation unless a new decision is made. Review analyses results, risks, and context. Renewal grants an additional period under updated conditions. Confusing the three allows ceremonial evaluations that change nothing. A committee may review a programme every year while it continues automatically. An effective clause, by contrast, requires evidence before the deadline and assigns someone the authority to retain, modify, or close the measure.
The decision should be published in understandable language: purpose, scope, duration, information used, results, observed harms, changes, and route of appeal. It should also state who will pay for transition. Exit costs —exporting data, training teams, informing users, restoring spaces, or temporarily maintaining two channels— belong in the original budget. If they emerge only at the end, economic dependence becomes an artificial reason to continue.
Institutional memory prevents the temporary from becoming invisible#
Organisations change staff and communities replace representatives. A measure can outlive the people who knew its origin. It therefore needs a living record that preserves the initial decision, assumptions, exceptions, contracts, complaints, and dates. Memory is not a passive archive: it enables a newcomer to ask whether circumstances remain the same and find the person who must answer. It also prevents the supplier from becoming the sole holder of technical knowledge.
Minor incidents deserve a record because they anticipate larger failures. An improvised login, a family missing notices, a physical barrier, or a repeated manual task are not isolated anecdotes when they accumulate. Periodic review should turn them into verifiable changes with owners and deadlines. If a promised correction reappears in every report, renewal requires an especially demanding justification or a pause.
Four institutional frameworks for testing without becoming trapped#
The European Commission's better regulation guidance proposes defining problems, comparing options, assessing impacts, and reviewing results. The OECD's strategic foresight work encourages institutions to explore diverse futures rather than extend a single forecast. UNESCO's Futures Literacy treats imagination as a capacity for perceiving alternatives and making better decisions in the present. UN-Habitat's people-centred smart cities programme places rights, inclusion, and participation within the governance of urban technology. These four frameworks do not provide a single recipe, but they converge on a discipline: an experiment needs purpose, evidence, participation, and the capacity to correct.
A community can translate that discipline into a short document. Before starting, it states what it intends to learn, what it is not authorised to do, who may stop it, when it ends, and how harm will be repaired. During the test, it publishes understandable signals and listens to affected groups. At the end, it compares alternatives rather than only measuring the selected option. Institutional sources offer criteria; the concrete decision still belongs to the people who must live with its consequences.
Repair includes knowing how to close#
When a measure causes harm, repair is not limited to correcting a record. It may require refunding a payment, restoring access, removing a label, providing assistance, deleting data, or publicly acknowledging an error. It must also prevent recurrence. An expiry date creates a moment when the accountable owner cannot postpone those obligations indefinitely. If the harm is serious, suspension authority must exist before the scheduled review.
Responsible closure protects those who benefited from the measure. It gives advance notice, offers transition, preserves necessary documents, and prevents the ending from falling hardest on the group with the fewest resources. For a service, the alternative must deliver an equivalent outcome; for a space, it must preserve accessibility and safety; for data, it must define return, portability, and deletion. Closing well is a public capability, not a confession of failure.
A common future is renewed through decisions, not inertia#
Mature communities may retain a provisional solution, but they retain it for fresh reasons. They recognise that time creates users, interests, and dependencies, and therefore raise the standard of proof. Renewal ceases to be an administrative formality when it includes absent voices, compares current alternatives, and distributes authority to correct. Learning is then no longer held captive by the tool that produced it.
The future as a common practice requires institutions able to begin, observe, change, and end. A temporary solution deserves continuity if it retains a legitimate purpose, demonstrates broadly shared benefits, reduces harm, and accepts another genuine review. If it cannot meet those tests, age does not make it necessary. The community then recovers the essential ability to say that an exception has served its purpose and should close, or that the time has come to transform it, with public rules and stable responsibilities, into something deliberately different.
