# Crime is not moving in one direction

Created 1 November 2026. UK Facts editorial explainer.

Knife offences, fraud and the long view: why recorded crime and survey estimates must stay separate.

## Figures, dates and limitations

### KNIFE OFFENCES: DOWN.
- England and Wales · police records · ONS
- Around −6% · reported and recorded offences

### FRAUD: UP.
- England and Wales · Crime Survey estimates
- ONS reports +11% using unrounded estimates.

### THE LONG VIEW IS DIFFERENT.
- England and Wales · to year ending March 2026
- Excludes fraud and computer misuse.

## Sources
- https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/bulletins/crimeinenglandandwales/yearendingjune2026
- https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/datasets/crimeinenglandandwalesappendixtables

Full analysis: https://dev.uk-statpack.co.uk/crime

## Production and assurance
AI narration: Jeff (Essex / North London). Casting is for a natural regional delivery; accent does not establish social class or geographic coverage. No impersonation or personal endorsement is implied.

Animated bars use zero baselines and the displayed units. No incompatible series are added together. Scene data and the finished file checksum are retained. Survey/estimate caveats remain beside figures. Policy proposals and commitments are labelled separately.

Render, caption timing, data-source checks and visual sample review are builder checks, not independent analytical assurance or AQuA certification. No guaranteed policy effects, invented savings or verdicts are claimed. The “What can we do?” film discusses selected commitments, not a comprehensive appraisal of the government programme.

No full 30-year series is manufactured where compatible history is unavailable. The education film uses provisional 2025–26 exam-gap data; attendance and support-plan measures have their own earlier dates. The trust survey is historical 2023 evidence, not current approval of the government.