Understanding UK Crime Statistics: A Homebuyer's Guide

Where the figures come from

In England and Wales, crime figures come from police.uk. The 43 police forces covering England and Wales, plus British Transport Police, send it what they record each month. Each crime is sorted into one of 14 categories, such as burglary, vehicle crime, or violence and sexual offences. Check Local adds up the last 12 months for each area.

Two things about this data matter before you read any number. First, it is published monthly and runs about two months behind. Second, locations are deliberately blurred: each crime is moved to the nearest of a fixed set of map points, usually the middle of a street or a public place, chosen so that a point covers at least eight addresses. That protects victims, but it means a crime pinned to your road may have happened a few doors or a street away.

Scotland works differently

Police Scotland does not publish street-level crime. The Scottish Government's Recorded Crime in Scotland statistics give yearly totals for each council. So a Scottish area page shows its council's figures, labelled as such, with no street-level heat map or category breakdown. Scottish and English or Welsh figures are built in different ways, so be careful comparing an area on one side of the border with one on the other.

Compare rates, not totals

A bigger place has more crime simply because more people live there. To compare areas fairly, use the rate per 1,000 residents. A town of 50,000 with 5,000 crimes a year (100 per 1,000) has a higher rate than a city of 500,000 with 40,000 crimes (80 per 1,000), even though the city's total is eight times higher.

On Check Local, every area's safety page sets its rate against the rate for its country. See Leeds for an example.

Look at the type of crime

The headline rate mixes very different things. For someone choosing a home, burglary and vehicle crime usually say more about daily life on a residential street than shoplifting in the high street does. Anti-social behaviour is the largest category in many areas, and how it is recorded varies between forces.

Trends need enough crimes to mean anything

One month can be skewed by a single incident, so use 12-month totals and compare a year with the year before. Small areas have few crimes, so their year-on-year change can swing a long way by chance. Where an area's own numbers are too small to give a reliable trend, Check Local leans on its council's figures and says so on the page.

What the data can't tell you

• It only counts what is reported and recorded. The ONS says many crimes, including fraud and domestic abuse, are under-reported. Recorded crime is a measure of police workload as much as of crime.
• Recording practice differs between forces and changes over time. A jump can follow a change in how a force records crime, not a change in crime.
• A source can have gaps. Greater Manchester's street-level data on police.uk is currently incomplete, so its crime counts are far too low. Check Local doesn't score safety in the ten Greater Manchester councils and doesn't publish the undercounted rate. See Manchester for how that looks.
• Numbers don't capture how a place feels. Visit at different times of day, and talk to people who live there.