The Extent of ‘Weeding’ of Stolen Vehicle Records on the PNC and DVLA

Brief / Summary

Date: 5 January 2026

Purpose: Oversight, assurance, and policy review

Abbreviations:

DVLA	The Driver & Vehicle Licensing Agency
LoS	Lost or Stolen
PNC	Police National Computer
V5C	Vehicle Registration Certificate – a.k.a. ‘log book’ or registration document
VRM	Vehicle Registration Mark a.k.a. vehicle registration number

An explanation of how differences between police-recorded vehicle theft figures and DVLA stolen-vehicle notifications are calculated, including the formulas used, to avoid ambiguity or misinterpretation of percentages can be found within the text and here.

1. Purpose and Audience

This brief summarises the findings of the accompanying report (below), The Extent of Weeding, which examines the potential systemic under-recording of stolen vehicles arising from police processes governing the confirmation and deletion (“weeding”) of Lost or Stolen (LoS) vehicle markers on the Police National Computer (PNC).

The document is intended for:

Police forces and Professional Standards Departments
Oversight bodies (HMICFRS, OPCCs, ICO)
Policymakers (Home Office, NPCC, DVLA)

It is not an allegation of misconduct or criminal liability. It is an evidence-based request for authoritative determination and governance review.

2. Issue in Summary

When a vehicle is reported stolen, a LoS marker is placed on the PNC. If that marker is not confirmed within a defined period (six weeks), it is automatically deleted (weeded).

Only confirmed LoS markers are transmitted to the DVLA. The consequence is that from day one, a vehicle may:

be reported stolen,
not be recovered,
yet cease to appear as stolen at the DVLA.

3. Why This Matters

Where weeding occurs at 6 weeks:

vehicles may appear legitimate in PNC, ANPR, DVLA and provenance checks,
innocent purchasers and insurers may be exposed to loss,
recovery opportunities are reduced,
police recovery and clearance metrics may be distorted,
national vehicle-theft statistics may be understated.

The issue is therefore one of data integrity, public protection, and governance, not administrative error.

4. Key Findings (High Level)

A. Persistent data discrepancy

Analysis of available data shows a material and persistent gap between:

police-recorded vehicle thefts (Home Office / force data), and
stolen-vehicle notifications received by the DVLA.

This discrepancy:

remains after accounting for timing mismatches and edge cases,
appears consistent with systematic deletion of unconfirmed records.

B. Identified mechanism

The six-week automatic deletion of unconfirmed PNC LoS markers provides a plausible and scalable explanation for the discrepancy.

C. Gwent Police as a case study

Gwent Police data illustrates the issue clearly:

vehicle theft figures recorded by the force and Home Office materially exceed DVLA stolen-vehicle notifications,
assurances that the issue had been resolved were undermined by a further confirmed weeding event in November 2025,
internal monitoring safeguards were reduced or removed, with reliance on manual confirmation.

D. National indicators

When DVLA and Home Office data are compared across forces:

wide variation is observed,
several forces show discrepancies exceeding 100%,
the pattern is inconsistent with benign explanations alone.

5. What This document Does — and Does Not — Claim

This paper does not claim:

that all discrepancies are caused by weeding,
that DVLA data is inherently unreliable,
that individual officers act improperly.

It does claim:

that automated deletion of unconfirmed LoS markers is a credible, systemic risk,
that the scale of the discrepancy requires formal determination,
that continued reliance on current processes undermines confidence and safeguards.

6. What Is Required Now

Authoritative determination of cause
A national body (most appropriately the Home Office, with NPCC and DVLA involvement) should establish whether weeding is a primary driver of the discrepancy and quantify its effects.
Governance and assurance review
Forces should demonstrate:
- who owns PNC LoS data integrity,
- what monitoring exists,
- how failures are detected and corrected.
Policy reconsideration
If confirmation at creation is feasible, the necessity of six-week automated deletion should be reconsidered at source.

7. Core Conclusion

If the discrepancy identified is substantially driven by weeding, the problem is systemic rather than exceptional.

Absent reform, stolen vehicles will continue to fall out of national records, public confidence will be eroded, and victims, insurers, and innocent purchasers will remain exposed to avoidable harm.

Next subject – 2

Report sections:

Further reading – weeding, caveats & the latest DVLA/Police LoS data.