Methodology

• Crime: data.police.uk monthly street-level crime files, one per force per month, for the latest 12 months available (Aug 2025 – Jul 2026).
• Population: Office for National Statistics population estimates for police force areas, mid-2024, all ages.
• Boundaries: ONS Open Geography Police Force Areas (December 2024), generalised, simplified further for display.

All three contain public sector information licensed under the Open Government Licence v3.0. Street-level files are what each force uploads to data.police.uk. They are not the official police recorded crime statistics published by the Home Office and ONS, and totals differ from those.

rate = recorded crimes in the month ÷ resident population × 1,000

Every rate is for one calendar month and uses the mid-2024 resident population of the force area. The same population is used for all 12 months.

City of London has 15,111 residents but a far larger daytime population of workers and visitors, so its rate (48.2 per 1,000 in Jul 2026) is not comparable with other forces. It is shown on the map and in tables and is ranked, but it is left out of the “highest rate” figure and drawn clipped at the edge of the rank plot.

• Anti-social behaviour: published in the same files but these are incidents, not recorded crimes.
• British Transport Police: not a territorial force and has no resident population.
• Police Service of Northern Ireland: outside England and Wales.

That leaves the 43 territorial police forces of England and Wales.

A force-month with no file is missing, never zero. It has no rate, is drawn hatched grey on the map and as a gap in trend charts, and is left out of national figures and rankings. Force-months not published in this window:

Greater Manchester published no street-level crime data in any month of this window, so it always shows as “no data”. The data.police.uk changelog lists Greater Manchester Police under known issues as having no crime data available and records its crime data as not provided for each month; it does not give a reason, so none is given here.

Some files exist but hold only part of a month. A force-month whose total is below 50% of that force’s median month in this window is treated as incomplete and handled exactly like a missing month: no rate, hatched grey, and left out of national figures, rankings and month-over-month changes. Showing it would present a partial upload as a collapse in crime.

The threshold is deliberately low so that a real fall is unlikely to be removed. That means a month can be visibly low and still be kept. Kept force-months below 80% of the force’s median, which should be read with care:

The data.police.uk changelog lists Northumbria’s crime data for Feb 2026 as not provided, although the archive contains a file for that month, so it may also be partial.

National rates, averages and ranks use only the forces reporting that month, in both the crime count and the population. Dividing by the whole population of England and Wales while some forces are absent would understate the rate. Ranks therefore read “Nth of N reporting”, and N changes from month to month.

The set of reporting forces affects the national figure by itself. In Jul 2026 the national rate is 7.73 across 40 forces. Had Metropolitan Police (8.64 per 1,000) not reported, it would be 7.56: a change of −0.17 per 1,000 with no change in crime anywhere.

Because of that, month-over-month changes compare only the forces reporting in both months, over the same combined population. A force that starts or stops reporting cannot move the change. The number of forces compared is shown next to the figure. On a force page, the change is that force against its own previous month, and is blank when either month has no data.

Rates are per calendar month and are not adjusted for month length. A 28-day February with exactly the same crimes per day as a 31-day January has a rate 9.7% lower, and the March after it a rate 10.7% higher. That part of the fall into February and of the rise out of it is the calendar, not crime. In this data, Feb 2026 is −4.3% on Jan 2026 like-for-like, but +5.9% per day; Mar 2026 is +13.9% on Feb 2026, but +2.9% per day. The same applies, less strongly, between 30- and 31-day months.

The map uses the same 7 bands of crimes per 1,000 residents in every month, so a colour means the same thing whichever month is selected. Each band includes its lower bound.

• <5.5
• 5.5–6.5
• 6.5–7.5
• 7.5–8.5
• 8.5–9.5
• 9.5–10.5
• ≥10.5
• No data