UK Crime

1. What is it?

UK Crime is a dataset that provides comprehensive crime risk information across the UK, offering insights into crime patterns by location. This data is derived from multiple authoritative sources, including police-recorded crime data and public health indices, such as the Scottish Index of Multiple Deprivation (SIMD). The dataset is designed to help users understand the relative crime risks associated with specific areas or properties, making it a valuable tool for insurance underwriters and pricing experts in evaluating and managing risk.

2. How can I use it?

Insurance underwriters and pricing experts can use UK Crime data to assess the level of crime risk associated with a specific address, neighborhood, or broader region. The data can be integrated into underwriting and pricing models to enhance risk assessment, adjust premiums, or modify policy terms based on localized crime risks. By leveraging this data, insurers can make more informed decisions, improve risk selection, and offer more competitive and accurately priced insurance products tailored to the actual crime exposure of each location.

3. How is the data created?

The UK Crime data is created by aggregating crime reports from multiple sources. The primary data source is UK Police data, which provides detailed information on reported crimes from police forces across England, Wales, and Northern Ireland. For Scotland, additional data is sourced from the Scottish Index of Multiple Deprivation (SIMD)%20into%20a%20single%20index.), which combines crime data with other socioeconomic factors to create a comprehensive crime risk profile. These datasets are processed, cleaned, and standardized to deliver consistent crime risk information that can be used for location-based analysis.

4. Are there any limitations?

UK Crime data is based on recorded and reported crime incidents, which means it is subject to the limitations of the underlying reporting systems, such as potential underreporting or inconsistencies in crime categorization. The data reflects historical trends and may not represent real-time crime conditions due to delays in reporting and data processing. Therefore, while UK Crime provides a valuable indication of crime risk, it should be used in conjunction with other data sources and contextual information for comprehensive risk evaluations.

5. What data points are available?

The UK Crime dataset includes the following attributes, each providing insights into different aspects of crime risk:

Burglaries by Postcode Sector (ukburg)

burg_score: A numeric score representing the risk level of burglaries in a given postcode sector.

burg_class: A numeric classification indicating the relative severity of burglary risk compared to other areas.

burg_trend: A numeric value showing the trend in burglary rates over time, indicating whether incidents are increasing, decreasing, or stable.

burg_bucket: A numeric value showing a simple bucketed score when ranking data nationally

Criminal Damage & Arson by Postcode Sector (ukcrda)

crda_score: A numeric score quantifying the risk of criminal damage and arson within a postcode sector.

crda_class: A numeric classification that categorizes the severity of criminal damage and arson risk compared to other areas.

crda_trend: A numeric value indicating the trend in criminal damage and arson incidents over time, reflecting whether these crimes are rising, falling, or stable.

crda_bucket: A numeric value showing a simple bucketed score when ranking data nationally

Vehicle Crime by Postcode Sector (ukvehc)

vehc_score: A numeric score assessing the risk level of vehicle crime in a specific postcode sector.

vehc_class: A numeric classification indicating the relative severity of vehicle crime compared to other areas.

vehc_trend: A numeric trend value showing the direction of vehicle crime rates over time, highlighting whether incidents are on the rise, declining, or remaining consistent.

vehc_bucket: A numeric value showing a simple bucketed score when ranking data nationally

6. Where can I find out more?

For more detailed information about UK Crime data you can explore the data sources listed above, our API documentation or contact our helpdesk using the chat option on this page.

7. Crime model Update April 2024

We’ve made several key updates and improvements to the UK Crime dataset.

The original 1 - 5 classification users currently consult remains the same, albeit with new data and some data gaps filled, as described below. As part of this April update we have take the opportunity to provide a new more granular 1 - 50 score.

Latest Data

• Updated to include the latest police crime statistics for the year ending 2024.

Coverage Fixes

• Resolved issues with data reporting from Northern Ireland and Greater Manchester, which were previously not providing complete crime statistics.

Updated Scoring Methodology

• Crime scores are now calculated using the latest full year of data available for each police force.
• We continue to provide trend fields covering a five-year period where data is available.

Enhanced _score Fields

• Introduced a refined 1–50 scoring distribution with more consistent and meaningful bands.
• This scoring is designed to support insurers seeking granular and standardised risk scores.

Continued Support for 1–5 Classes

• The traditional 1–5 classification is still available.
• Based on a normal distribution:
• Most properties will score a 3.
• Properties in high or low-risk areas will score 1, 2, 4, or 5.

New _bucket Field

• The previous “score” values (simple buckets) are now available in the new _bucket field.
• Ideal for customers who need to reference or migrate from the old classification system.

New Scoring Bands

Below are the new approximate scoring bands for the 1 - 50 distribution. Note that these bands are at LSOA level and so when mapped to specific postcodes or properties these may vary as LSOAs are not equal in size. Users can contact their Addresscloud account representative for a spreadsheet showing distributions of postcodes against each of these scores.