Methodology

Sources

• Home Office: Police misconduct, England and Wales (annual statistical release).
• Independent Office for Police Conduct (IOPC): complaints and review statistics.
• HMICFRS: PEEL inspection grades, thematic reports, and Super-Complaint responses.
• UK Parliament: bills, written and oral questions, Hansard, committee output, consultations and calls for evidence (via the official open APIs).
• WhatDoTheyKnow (mySociety): FOI request search.
• Office for National Statistics: crime, workforce and population data used for rates.
• Force-published responses to Super-Complaints (CWJ, etc.).
• HMICFRS Policing Performance System: force performance levels, assigned three times a year, drawing on inspection findings and on Home Office Police Performance Framework data among other inputs.

Definitions

Misconduct.

Conduct that, if proven, would justify disciplinary action under the Police (Conduct) Regulations 2020.

Gross misconduct.

Conduct so serious that dismissal without notice could be justified.

Upheld.

An allegation found proved at a misconduct meeting or hearing.

Sanction.

The outcome imposed: written warning, final written warning, reduction in rank, dismissal with or without notice.

PEEL grade.

HMICFRS grade in one of five bands: outstanding, good, adequate, requires improvement, inadequate.

Performance level.

The level HMICFRS assigns a force under the Policing Performance System, from Level One (the default, no intervention required) to Level Four (special measures). A level is not a grade. Level One is the default position and does not indicate the absence of performance issues.

Refresh cadence

Static datasets (misconduct, IOPC complaints, PEEL grades) are refreshed when the source publishes a new release. Live feeds (Parliament, HMICFRS publications, FOI search) call the source API at view time, with a short cache. Every monitor and every force page carries a last updated timestamp. Performance levels are reassigned three times a year at the Policing Performance Monitoring Group, with the next round expected in autumn 2026. We record the date each level was assigned.

Normalisation

Comparisons across forces use rates per 1,000 officers or per 100,000 residents, not raw counts, because force sizes range from about 1,000 to over 30,000 officers. Where the underlying release uses a different denominator we say so on the page.

Limitations

• Misconduct statistics under-count. Many allegations are handled locally and never enter the formal pipeline.
• Complaint data depends on whether a complaint was recorded - a known weakness in several forces.
• PEEL grades reflect the inspection at the time and are not real-time performance measures.
• A Level One assignment is the default position rather than a positive assessment. A force can hold Level One while carrying open areas for improvement from its most recent PEEL inspection.
• Performance levels are a snapshot at the date of assignment and cover the force as a whole. A force at Level One may still carry open areas for improvement or a requires improvement PEEL grade.
• Free-text source pages can change. We link to canonical URLs and timestamp our extracts.

Corrections

If a figure on the site is wrong or out of date, please tell us. We treat corrections as a priority. See About this site for how to get in touch.

Citing this site

Victim Not Suspect, victimnotsuspect.org, accessed [date]. Methodology: victimnotsuspect.org/methodology.