How the crime comparisons work

The official series

• Counts describe police-recorded crime, rather than every crime that occurs. Reporting levels, recording practices and police activity affect comparisons and trends.
• Each count covers a rolling 12-month period. Year-on-year change compares the count with the preceding rolling year ending in the same quarter.
• Fraud and anti-social behaviour incidents are excluded. Weapons possession is not the same measure as knife-enabled crime.
• Rates are calculated from exact ONS mid-year resident-population estimates, not inferred from rounded published rates. Historical rates use the latest available estimate at or before the year preceding each period end. Historical population estimates may be revised.
• Standard crime histories sum four consecutive quarters from the latest Home Office revised open-data file. Latest and previous-year counts are reconciled against ONS Tables P1 and P4, and published percentage changes/suppressions against P2, before publication. Knife totals reconcile between P5 and P6 and current rates against P7; firearm counts and their dated histories come directly from P8.
• City of London counts remain visible, but its resident-based rates and rate rankings are suppressed. London regional totals include City of London.
• This version covers England and Wales. Crime recording systems and offence definitions differ in Scotland and Northern Ireland.

Rates, coverage and overlap

A rate divides offences by the resident population, then multiplies by 1,000 or 100,000 as stated. Regional rates divide total offences by total residents; they are not the average of force rates. Visitors and commuters can affect comparisons. The City of London remains visible by count but is not ranked by resident rate.

• The England and Wales reference includes British Transport Police except for firearm offences. Regional and territorial-force comparisons exclude it. London regional totals include City of London.
• Main offence groups and their detailed components overlap. Knife-enabled and firearm offences also overlap other categories. Never add parent groups, their components and weapon-enabled measures together.
• Published ONS percentage-change suppressions are preserved, including small-base suppressions. Some observations have counts but no published comparable change.
• Standard offence histories contain quarterly rolling-year totals. Knife-enabled and firearm histories retain the actual published annual dates and latest equivalent rolling years; intervening quarterly observations are not estimated. Historical weapon rates are unavailable before a matching ONS population estimate exists.
• An April 2025 clarification of police counting rules increased the recording of business robbery during shop theft involving violence or threats. Changes in robbery and shoplifting across this date can reflect the recording change as well as changes in crime.

The separate monthly series

• Monthly provisional Police.uk records. Broad categories differ from the official quarterly series. Areas with missing submissions are excluded from the latest comparison; missing data never becomes zero.
• Rates use the same ONS resident-population estimates and police-force geography as the official comparison. They are not a resident's probability of becoming a victim.
• Each current value is a complete rolling 12-month total. Change compares it with the preceding 12 months. History shows rolling-year totals ending each month; individual monthly counts are included in the data export.
• Violence and sexual offences are combined by Police.uk. Fraud and anti-social behaviour are excluded. ASB incidents must not be interpreted as criminal offences.
• A police counting clarification from April 2025 treats shop theft involving violence or threats, including during escape, as business robbery. Robbery and shoplifting trends span this recording change.
• Every CSV row is counted by its reporting force, including records with missing coordinates. British Transport Police and Northern Ireland are excluded. Regional totals sum constituent territorial forces.
• A missing force-month CSV makes the corresponding rolling-year count unavailable for that force and every containing region or national total. Presence of a CSV does not guarantee perfect reporting: consult Police.uk's changelog.
• Police.uk publishes snapshots which may not incorporate later crime cancellations and reclassifications unless the force refreshes its data. Compare trends within this source; do not splice it into the official quarterly time series.