A Consultation Is Not a Referendum: Write a Response That Can Be Used

Featured image: conceptual AI-generated illustration created for this article; it is not documentary evidence.

A public consultation can look like a vote. There is a proposal, a deadline and a box asking whether you agree. That resemblance creates a dangerous misunderstanding: if 10,000 people select “no”, many assume the policy must be withdrawn; if only a handful reply, others assume their contribution cannot matter.

Consultations generally do a different job. They help a public body test proposals, understand consequences, discover missing evidence and improve implementation before a decision is finalised. Numbers may matter, but responses are not necessarily counted like ballot papers. One well-supported submission can reveal a legal conflict, an operational cost or an affected group the proposal overlooked.

That does not make consultation automatically fair or influential. The scope may be narrow, the questions may be technical, and decision-makers may proceed despite widespread disagreement. The practical task for a respondent is to identify what is genuinely open to influence and submit information in a form that can be examined.

First identify the democratic instrument

A consultation is not the same as a referendum, election, petition, survey or parliamentary call for evidence.

• Election or referendum: eligible votes determine representation or a specified public choice under formal rules.
• Petition: people collectively ask an institution to take or consider an action; thresholds may trigger a response or debate.
• Survey: gathers information from a sample or group, sometimes for research rather than a formal policy decision.
• Government consultation: seeks views and evidence on a proposal, policy or implementation question.
• Parliamentary call for evidence: invites information that a committee can use to scrutinise an issue and develop findings or recommendations.

These routes can overlap, but their rules and consequences differ. Before writing, confirm who is asking, which power they hold, the decision stage and what they promise to do with responses.

Read the scope before arguing the case

Begin with the consultation page, main document, impact assessment and response instructions. Extract five things:

1. the proposal being considered;
2. the questions on which views are requested;
3. the matters already decided or outside scope;
4. the evidence or options used to justify the proposal;
5. the closing date and accepted submission method.

The UK Government’s Consultation Principles say consultations should be clear and concise, have a purpose, be informative and occur when plans are still at a formative stage. The guidance also says timing should be proportionate and that departments should publish responses within 12 weeks of closing or explain why this is not possible.

Those are principles for government departments, not a guarantee that every consultation will feel accessible or produce the outcome a respondent wants. They do, however, provide a useful test. If the document is unclear about what can change, say so precisely and explain how that ambiguity limits meaningful engagement.

A consultation is usually not a headcount

Some exercises publish how many respondents supported or opposed an option. That statistic can describe participation, but it does not automatically operate as a binding vote. Respondents may be self-selecting; organised campaigns may send similar messages; and submissions may contain very different levels of evidence.

The Crown Prosecution Service makes the distinction unusually explicit in its consultation principles: each response is considered on its merits, and a single sufficiently strong response may lead to an amendment; the quantity is noted but does not determine the merit of changing guidance.

Different public bodies and consultations may use different methods, so read the stated process. The broader lesson is that “many people said this” and “the evidence supports this” are separate claims. A strong response can contain both, but should not confuse them.

Build the response around claims and consequences

For every important point, use a four-part structure:

1. Position: State whether you agree, disagree or support the proposal only under certain conditions.
2. Reason: Explain the mechanism, not merely the emotion.
3. Evidence: Provide verifiable data, research, professional knowledge or clearly labelled lived experience.
4. Recommendation: Describe the change you want and the likely effect.

Compare two responses:

“This change is unfair and will hurt families.”

“I oppose the proposed two-week application window because carers in our service often need longer to obtain medical evidence. In our 2025 records, 31 of 84 completed applications required more than 14 days. I recommend a 28-day window with a documented extension route for evidence outside the applicant’s control.”

The first statement may be sincere and politically important. The second gives an analyst a specific mechanism, a sourceable observation and a feasible amendment. If the numbers are from an organisation’s records, explain the sample, period and limitations. Do not present anecdote as a population estimate.

Lived experience is evidence—label it accurately

People sometimes apologise for not being policy experts. Yet consultations can fail precisely because formal analysis misses how systems work in daily life. A resident may know that an “accessible” office sits behind a broken lift. A parent may understand that a proposed digital form assumes literacy or technology that families do not have. A worker may recognise that an official process is routinely bypassed under time pressure.

Present lived experience with useful context:

• what happened;
• how often or over what period;
• who was affected, without exposing unnecessary personal data;
• which part of the proposal would improve or worsen the situation;
• what is observed directly and what is your interpretation.

One account does not prove prevalence, but it can demonstrate possibility, expose a mechanism or suggest a question for further investigation.

Answer the question—but record what the question misses

Consultation questions can force complex views into “yes”, “no” or “not sure”. Select the closest answer, then use the explanation field to state conditions and exceptions.

If a crucial issue is absent, do not abandon the requested structure. Answer the listed questions first, then add a clearly labelled section such as “Material issue not covered by the questions”. Explain why the issue falls within the consultation’s purpose and what evidence supports its inclusion.

This makes the response easier to code and analyse while preserving legitimate challenge. A long essay that ignores the actual questions may be emotionally satisfying but harder for the consultation team to use.

Separate evidence, interpretation and preference

A credible submission distinguishes three kinds of statement:

• Evidence: “The published impact assessment estimates implementation costs of £4 million.”
• Interpretation: “The estimate may understate training costs because it excludes temporary staff.”
• Value judgement: “Reducing exclusion is worth a modest increase in administration.”

Politics involves values; there is no need to disguise them as neutral facts. The discipline is to show the reader which part of the argument is observed, inferred or preferred.

Use sources that can be checked

Link directly to the relevant table, report or statutory provision where possible. Give the title, author, publication date and page or section. For original organisational data, explain how it was collected. Remove personal identifiers unless disclosure is necessary, lawful and explicitly accepted by the process.

Avoid burying the main point beneath attachments. Summarise what each source demonstrates and why it matters. An evidence dump transfers the analytical burden to the reader.

Make recommendations proportional and implementable

“Do not proceed” can be a legitimate recommendation, especially where the harm cannot be mitigated. But many submissions become more useful when they also offer alternatives:

• change a definition;
• add an exemption or safeguard;
• pilot the proposal before national rollout;
• extend a deadline;
• publish disaggregated monitoring data;
• require independent review;
• include a sunset clause or formal review date;
• provide a non-digital route.

State who should act, what should change and how success or harm could be measured.

When Parliament asks for evidence

Select committee inquiries are different from departmental consultations. Committees scrutinise policy, administration and spending, hear evidence and usually publish reports. The UK Parliament’s written-evidence guidance advises contributors to read the call carefully, keep submissions short and focused, use clear headings and numbered paragraphs, and concentrate on the expertise or experience they can add.

A submission may be published and attract parliamentary privilege, but contributors should read the rules on confidentiality, personal data and material already published elsewhere. Do not assume that sending a document to a government department also submits it to a parliamentary committee; use the specified portal or address.

Check what happened afterwards

Save the consultation page, your submission and confirmation receipt. After closure, look for:

• a summary of responses;
• the government or public body’s formal response;
• revised proposals or draft legislation;
• an impact assessment;
• an explanation of rejected recommendations;
• implementation and review arrangements.

Compare the outcome with the original proposal. Did the body identify themes, explain changes and address contrary evidence? A consultation should not be judged only by whether your preferred outcome won. It should also be judged by whether the process showed how relevant evidence was considered.

A practical submission checklist

1. Confirm who is consulting and what authority they hold.
2. Write down what is open to change and what is outside scope.
3. Read the proposal, impact material and response rules.
4. Answer the actual questions in their stated order.
5. For each key claim, add reason, evidence and recommendation.
6. Label lived experience, organisational data and inference accurately.
7. Use verifiable links and explain what each source shows.
8. State trade-offs and who may bear costs or risks.
9. Remove unnecessary personal and confidential information.
10. Submit before the deadline and retain proof.
11. Track the formal outcome and any implementation review.

Influence is more than volume

Collective mobilisation matters in politics. A large response can show intensity, organisation and public concern. But a consultation response has another route to influence: it can make a decision-maker’s model of the problem harder to ignore.

The strongest submission does not merely announce a side. It identifies a consequence, supports it honestly, proposes an action and leaves a trail that others can scrutinise. A consultation is not a referendum—but it is still an opportunity to put usable knowledge into the public record.

Primary guidance and further reading

• UK Government: Consultation Principles 2018.
• UK Parliament: Giving written evidence to a select committee.
• UK Parliament: Current committee inquiries and calls for evidence.
• Crown Prosecution Service: Prosecution guidance consultation principles.