Crime context - worksite in durham

This worksite

The worksite buffer spans 2 LSOAs - context for each is below.

County Durham 062A (E01020825)
Population (2021 Census): 1,733
Safety-relevant crimes (a rolling 12-month window (2025-06 to 2026-05)): 148
Adjusted rate: 64.1 / 1,000 residents / year
Band: above force typical

By category:

Violence and sexual offences: 81
Anti-social behaviour: 49
Public order: 15
Possession of weapons: 2
Robbery: 1

County Durham 063B (E01020826)
Population (2021 Census): 1,500
Safety-relevant crimes (a rolling 12-month window (2025-06 to 2026-05)): 113
Adjusted rate: 57.5 / 1,000 residents / year
Band: above force typical

By category:

Violence and sexual offences: 68
Anti-social behaviour: 39
Public order: 6

What this page is

Recorded-crime context for one worksite, to inform a lone-worker deployment decision - two-person crew, check-in interval, lighting, vehicle positioning - not a yes/no on lone working, and not a risk score.

Method

Counts are a rolling 12-month window (2025-06 to 2026-05) of recorded crime in the categories that bear on a lone worker's confrontation/threat/assault risk (violent crime, public order, anti-social behaviour, robbery, possession of weapons) - not property crime.
The window always ends at the most recent month data.police.uk itself confirms it has published, never a fixed guess back from today.
The denominator is each LSOA's 2021 Census usual-resident population, not land area - LSOAs are population-equal by design, so a rate per 1,000 residents says something an area-based rate cannot.
Rates are shrunk toward the force-wide average, more so for smaller populations, then compared in quintiles (or terciles below a floor, or not banded at all below a lower one) within this force's own LSOAs only.
Grey areas had too few recorded crimes over the window to compare meaningfully - not "zero crime", just not enough data to say anything.

What this data cannot tell you

There is no time of day in this data. It is month-level only. An area with high daytime public-order counts near a shopping precinct is a different proposition at 2am than the same count near a pub strip at closing time - the source cannot distinguish them, and this is probably the single most important limitation for a lone-worker decision.
Anti-social behaviour is reporting-sensitive. It is likely the most relevant category here for abuse directed at road crews, and it also reflects how readily an area's residents report things as much as what actually occurs - it should not be the swing factor on its own.
Each LSOA is derived from crime locations already snapped to anonymised points, so it inherits that displacement - tolerable at LSOA size, least reliable for a worksite sitting near an LSOA's edge.
No per-street or per-address figure is reported: a real map point covers at least eight postal addresses, or none, and the true coordinates are already replaced by the map point's - LSOA sits above that floor, nothing finer does.
No comparison across forces or nationally - bands are quintiles within this force's own LSOAs only.
LSOA boundaries are far more stable than the neighbourhood-policing boundaries the predecessor to this example used, but are not literally fixed; this aggregates the window's crime against today's boundary.

This is one input among several a deployment decision should weigh: sightlines, egress, proximity to licensed premises, whether the crew works in hi-vis beside a lit vehicle or on foot after dark. It informs; it is not a lone-worker risk assessment and is not a substitute for one.