England Planning Appeals Monitor

Read England Planning Inspectorate appeal decisions from the monthly Casework Database. Filter cases by local authority, casework type, development type, decision, date, and residence count. Results appear in newest decision date order.

The actor discovers workbook attachments through the GOV.UK Content API. It streams the Data Set sheet instead of loading the workbook into memory. The default run reads the current Casework Database and emits up to 100 decisions from the last 90 days.

Run locally

Use Node.js 20 or later.

npm install
npm test
npm start

An empty input contacts the live source. Tests use local fixtures and stub network calls. Read results in storage/datasets/default and the summary in storage/key_value_stores/default/RUN_SUMMARY.json.

Input

Field	Default	Meaning
lpaNames	[]	Case-insensitive substrings of local planning authority names.
onsCodes	[]	Case-insensitive exact matches against either authority ONS code.
caseworkTypes	[]	Case-insensitive exact casework type matches.
developmentTypes	[]	Case-insensitive exact development type matches.
decisions	[]	Case-insensitive exact decision matches.
decidedAfter	90 days before the run date	Exclusive decision date lower bound, in YYYY-MM-DD format.
minResidences	Unset	Minimum residence count, including zero.
onlyNew	false	Set true to emit only new and changed cases.
maxItems	100	Maximum emitted records, from 1 to 10000.
includeOlderCasework	false	Also read the Older Casework Data attachment.
stateStoreName	england-planning-appeals-monitor-state	Named key-value store for case history.

List filters accept arrays of strings. An empty array applies no filter. Values within one array are alternatives. Different filter categories all apply. Records without a decision date do not match the decision date window. Records without a residence count do not match minResidences when it is set.

includeOlderCasework does not widen the date window. Set an earlier decidedAfter to retrieve older decisions. maxItems limits emitted results. The actor must scan the selected workbook attachments to find the newest matching decisions.

Output

This illustrative record shows the output shape. It is not a captured appeal.

{
  "caseNumber": "1234567",
  "caseworkType": "Householder (HAS)",
  "lpaName": "Example Council",
  "lpaOnsCode": "E00000000",
  "localAuthorityOnsCode": "E00000000",
  "lpaDecisionDate": "2026-06-01",
  "receivedDate": "2026-06-10",
  "validDate": "2026-06-12",
  "startDate": "2026-06-15",
  "decisionDate": "2026-10-01",
  "decision": "Dismissed",
  "procedure": "Written Representations",
  "developmentType": null,
  "appealTypeReason": null,
  "appealReason": null,
  "typeDetail": null,
  "jurisdiction": "England",
  "linkStatus": null,
  "leadCase": null,
  "appellantType": "individual_or_unknown",
  "siteAddress": null,
  "sitePostcodeDistrict": "HA9",
  "areaHectares": null,
  "floorSpaceSqm": null,
  "numberOfResidences": null,
  "lpaApplicationReference": null,
  "greenBelt": null,
  "costsApplied": null,
  "enforcementGrounds": null,
  "appealUrl": "https://acp.planninginspectorate.gov.uk/ViewCase.aspx?caseid=1234567",
  "source": "Planning Inspectorate Casework Database, https://www.gov.uk/government/publications/planning-inspectorate-appeals-database",
  "licence": "Open Government Licence v3.0, http://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/",
  "retrievedAt": "2026-10-22T00:00:00.000Z",
  "generatedId": "1234567",
  "changeType": "new"
}

Dates are ISO calendar dates. Missing source values are null. generatedId is the case number. changeType is new, changed, or unchanged.

The actor compares a hash of the emitted source fields with the saved hash for the same generatedId. Decision, date, and other output changes can mark a case as changed. Retrieval timestamps and change metadata do not affect the hash. State saves only after a successful result push. A crash between the push and state save can repeat a result on the next run.

Use a separate stateStoreName for independent histories. Change detection covers cases that pass the current filters and are emitted within maxItems. It is not a complete audit of every historical case.

RUN_SUMMARY records run counts, timestamps, and any run error. The actor charges per emitted result using the result event, with no run fee. Unchanged results emitted with onlyNew=false also count as results.

Personal data

The output never contains the source Appellant, Agent, or Inspector Name fields. The actor uses Appellant only to classify appellantType with the organisation pattern specified for this project. An organisation match is a heuristic, not a verified legal status.

A full siteAddress appears only for an organisation match outside these casework types:

• Householder (HAS)
• Enforcement Notice
• Lawful Development Certificate
• High Hedges
• Tree Preservation Order
• Tree Replacement Notice
• Appeal against Enforcement Listed Building Notice/Enforcement Conservation Area Notice

All other cases have siteAddress: null. sitePostcodeDistrict contains only the postcode outward code, such as HA9, when the address contains a recognised UK postcode. The actor does not produce inspector-level statistics.

Source credit and licence

Contains public sector information licensed under the Open Government Licence v3.0. Source: Planning Inspectorate Casework Database. The actor normalizes dates, filters records, removes personal fields, and applies the address restrictions above. No endorsement by the Planning Inspectorate is implied.

The England source and licence text below is reproduced from TERMS.md.

• Source: Planning Inspectorate Casework Database, https://www.gov.uk/government/publications/planning-inspectorate-appeals-database (updated 2026-10-01, monthly xlsx, rolling 5 years). Attachment URL read from the GOV.UK Content API https://www.gov.uk/api/content/government/publications/planning-inspectorate-appeals-database.
• Licence: GOV.UK content under OGL v3 unless stated; the same dataset on https://opendatacommunities.org/data/planning-inspectorate-casework and data.gov.uk lists OGL v3 (http://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/). OGL v3 does not cover personal data. Data notes (https://www.gov.uk/government/publications/planning-inspectorate-appeals-database/appeals-data-data-notes-and-guidance): third-party copyright needs permission; data can change between publications; PINS states patterns by Decision Maker are artefacts and not evidence of fitness.
• robots: gov.uk robots.txt disallows only /search/all and print pages.
• Conditions and how met: OGL attribution in README and items; Appellant, Agent and Inspector Name are never output (only appellantType = organisation/individual/unknown from a heuristic, no names); full site address is output only when the appellant name matches an organisation pattern (Ltd, Limited, LLP, plc, Council, Homes, Developments, Group, Trust, etc.) and the casework type is not Householder (HAS), Enforcement Notice, Lawful Development Certificate, High Hedges, Tree Preservation Order or Tree Replacement Notice; otherwise only the postcode outward code (e.g. HA9) is output. No inspector-level stats.

Limitations

The source updates monthly, so decisions and corrections can arrive after their decision dates. The current workbook covers a rolling five-year period. Source values can change between publications. A short date window does not detect revisions to older decisions.

The actor downloads the selected workbooks on each run. Workbook size and source availability affect run time. It streams rows and retains matching output records rather than workbook cells. Postcode extraction and organisation classification can miss unusual formats. The public appeal link may not expose every case detail.

Verification with the source sample

samples/ is excluded from git. With the supplied samples/full.xlsx, run node scripts/verify-sample.js to scan the full workbook through the actor flow using a file-backed test adapter. It writes 100 default-filter records and RUN_SUMMARY under storage/sample-verification, reads them back, checks privacy and ordering, and checks peak process memory against 1 GiB. This verifies the sample pipeline without contacting GOV.UK or loading the Apify SDK.

Run python3 scripts/build-fixture.py to rebuild the trimmed workbook fixture from the sample. The script replaces names and site addresses with test values before writing the fixture.