Reading a drafted report for unsupported claims

Unsupported claims in a planning report are the ones that sound reasonable and cannot be backed. A sentence saying the council usually asks for a parking survey on schemes of this size. A figure with no source. A conclusion that no protected species are present when the only fieldwork was a two-hour walkover. Each one is a place where a case officer, a consultee or an inspector can push, and on a contested application they will.

This is an editorial pass, separate from checking that policy citations are correct. You are looking for assertions, not errors.

Claims about council behaviour

Anything of the form the council normally, this authority typically, officers here usually, or they will probably accept is an assertion about a third party that you cannot evidence.

Some of these claims are true. Practitioners who work in an area know what a validation team asks for. The problem is that the knowledge is tacit and the document is public, and a statement about how an authority behaves can be turned round on you if the authority behaves differently this time.

Three fixes, in order of preference:

• Replace with a citation. The local validation checklist, the adopted local plan policy, the supplementary planning document, a published committee report, or the council's own pre-application charging page. Name the document and its date.
• Replace with a factual statement about your own case. Rather than the council usually requires a transport statement, write the pre-application response dated 14 May confirmed that a transport statement would be required.
• Delete it. If neither of the above is available, the sentence was decoration.

The same rule applies to statements about consultees. National Highways, the lead local flood authority and the county archaeologist all publish standing advice, and citing it is stronger than describing what they tend to say.

Figures without a stated source

Every number in a planning document should have one of three things behind it: a cited source, a calculation you can show, or a measurement you took.

Work through the numbers in this order, because it is roughly the order of risk:

1. Percentages. A percentage without a numerator and a denominator is the most common invented figure in planning writing.
2. Costs and values. Use a published fee schedule, a charging schedule, or a quoted figure with its date and provenance.
3. Distances and areas. Measure them and say from what: the site boundary, the nearest habitable window, the edge of the designation.
4. Timescales. Statutory periods are fine and citable. Estimates of how long an authority will take are not.
5. Survey and count data. Attribute to the survey, with its date and method.
6. Trip rates, occupancy rates, population figures. Name the database or the census year.

Where a drafted report has inserted a marker because a site fact was missing, that marker is doing its job. Resolve it with a number you can stand behind, or rewrite the sentence so it does not need one. Deleting the marker and leaving a vague phrase is the worst of the three options.

If a number cannot be sourced, describe the effect without it. The scheme increases the built footprint is defensible. The scheme increases the built footprint by around 40 per cent is not, unless you did the arithmetic and can show it.

Conclusions that outrun the survey

This is the category with the most professional risk attached, and it is where drafting tools and junior staff make the same mistake.

A preliminary ecological appraisal supports a statement that no evidence of roosting bats was recorded during the daytime inspection on a given date, and that further survey is recommended. It does not support a statement that bats are absent. Absence needs emergence surveys in season, carried out by a suitably licensed ecologist, and protected species work is legally required to be done by one.

The pattern repeats across disciplines:

The fix is to name the survey, its date, its method and its limitations in the same sentence as the conclusion. It reads as more cautious. It is also much harder to attack.

Where the report is an environmental statement chapter, the competence question is statutory as well as editorial, as our note on competent experts and the EIA Regulations 2017 sets out.

Comparisons with other sites

Comparative arguments are useful and they need references.

A statement that similar schemes have been approved in the area is worth nothing without application numbers. Give the reference, the address, the decision date and one sentence on why the case is comparable. Then give the material differences too, because the case officer will find them.

Appeal decisions are stronger evidence and still not precedent in the binding sense. Describe an inspector's decision as indicative of how the issue has been approached, cite the appeal reference and date, and quote the relevant passage rather than characterising it. An inspector reading your statement may well have the decision in front of them.

Watch for the comparison that quietly becomes a claim about outcome. Saying that a comparable scheme was approved is a fact. Saying that this scheme should therefore be approved is an argument, and it needs to be made as one, with the planning balance set out.

Statements about approval prospects

No document you submit should predict the decision. A planning statement argues that the proposal accords with the development plan and that material considerations support it. It does not say the application will be approved.

Three forms of this creep in. Direct predictions. Reassurances to the client that have survived into the submitted version. And phrases like fully compliant or meets all requirements, which are claims of certainty about someone else's judgement.

PlanScope does not guarantee planning approval or validation, and no drafted document should suggest otherwise. PlanScope provides AI-assisted draft reports for planning applications, and reports should be reviewed before submission.

The same discipline protects you commercially. A statement predicting approval, relied on by a client who then buys the site, is the kind of sentence that gets read aloud later.

Marking and resolving each unsupported claim

Use a consistent marker so the pass is mechanical rather than a matter of judgement each time.

Read once with a highlighter, or a comment tag, and mark every sentence that falls into one of the five categories above. Do not fix anything on this pass. Fixing interrupts the reading and you stop noticing.

Then work through the marks in a table: the sentence, the category, the action, and the owner. Actions are limited to four. Cite it, calculate it, qualify it, or cut it. Anything that cannot take one of those four is cut.

Assign an owner for each item that needs someone else's input, with a date. On a multi-author document this is the step that stops an unresolved claim reaching submission, and it dovetails with building an internal review checklist for submissions.

A review pass that takes twenty minutes

For a document of ordinary length, a search-driven pass is faster than reading and catches more.

1. Search for digits. Check each number against its source. Ten minutes on most documents.
2. Search for typically, usually, generally, often, normally, tends to and commonly. These words introduce most claims about council behaviour.
3. Search for all, no, none, never and always. Absolute statements rarely survive examination.
4. Search for approv to find every claim about outcome.
5. Search for similar, comparable and precedent to find comparative claims without references.
6. Search for the confirmation markers the draft inserted, and resolve every one.

That is the twenty minutes. Factual accuracy is a separate pass, covered in our note on factual errors to look for in a drafted planning document, and the authorship question is covered in when a council asks who wrote your report.

Where a document needs a qualified reviewer before it goes in, PlanScope's professional directory is the route to request a review and sign-off. For specialist technical reports, qualified professional input may be required.