Crime in London

  1. Home

Get help


This site helps you find out about crime in your neighbourhood and take action to keep yourself safe. This page includes answers to commonly-asked questions, but if you have any other questions then please contact me.

Who runs this site?

This site is run by Matt Ashby, a research student at the UCL Department of Security and Crime Science. I run this site because I am interested in is how police can work to reduce crime, and accurate information about what crime is happening is key to this. You can find more information about my research (which is mainly about transport crime). I tweet as @lesscrime.

Data sources

The crime data used on this site is from Demographic (census) data is from Neighbourhood Statistics and the London Datastore. These datasets are all licensed under the Open Government licence, which allows re-use of most UK government data.

At the start of each month, the Home Office collects crime data from all of the police forces in England and Wales, processes them and releases them on the website. The data processing takes about a month, so new data are usually released at the end of the month after the month that the data refer to. For example, crime data for January is usually released on the last day of February. After the data are released by the Home Office, I analyse them (most of this is done automatically) and then update the pages for each borough and neighbourhood.

Why doesn't Crime in London include all the data?

The Home Office releases 11 different categories of crime data, but Crime in London only includes detailed information for six of those categories. The other categories are not covered in detail because there is reason to believe that police-recorded crime statistics are not an accurate reflection of how many crimes actually happen. This means it could be misleading to analyse how the number of these offences changes over time, or to compare these offences in different areas, so I do not include detailed information on them.

The specific reasons for this are:


This category includes all the offences that cover making, selling, distributing or (much more commonly) possessing any controlled drug. These offences are almost always never reported to the police, so most recorded drug offences are the result of police officers discovering drugs on a person or in a building.

Since the police can only record drug offences when they discover them, the number of recorded drug crimes in a particular area is likely to reflect the number of police officers in that area looking for drugs, rather than accurately reflecting the number of drug offences going on in that area.

‘Public disorder and weapons’

This category includes includes offences when an offender verbally abuses or threatens a victim, but it also includes more serious (and much less common) breaches of public order such as riots. It also includes possession of weapons such as knives, although these offences are generally less common.

There are two reasons why the number of recorded crimes in this category is not an accurate reflection of how many crimes actually happen.

  • Possession of weapons (such as knives) is usually only recorded when police officers discover people carrying weapons, so (as is the case with drug crimes) the number of recorded weapons offences is likely to reflect the number of police officers patrolling in an area.
  • The Home Office does not release detailed information on how many of each type of crime makes up each category, but it is likely that most crimes in this category are offences against sections 4, 4A or 5 of the Public Order Act 1986. A very common example of this type of crime is a person shouting and swearing at someone else in the street.

    The Home Office sets strict rules for when the police should and shouldn't record a crime, which say that these incidents should not be recorded as a crime unless the disorder is still happening when police arrive or a specific victim is at the scene. This means that whether a crime is recorded or not depends partly on whether or not people think the incident is serious enough to report to the police, how long it takes for the police to arrive at the scene and whether or not the victim has waited at the scene for officers.

Together, this means that the number of recorded crimes in this category is more likely to be determined by the availability of police officers rather than on the number of crimes that happen.


This category includes all thefts from shops except burglaries and robberies. Detailed information about shoplifting is not provided on the Crime in London site because people are very reluctant to report shoplifting, so as little as 5% of shoplifting is recorded by police. Since reporting rates are likely to vary depending on the policies of individual shops (some shops pay bonuses to security staff who catch shoplifters, while others do not employ security guards at all), it could be misleading to compare rates of shoplifting across different areas.

‘Other crime’ and ‘other theft’

Detailed information about these categories is not included on the Crime in London site because the categories are too broad. The ‘other theft’ category includes snatches of mobile phones on the street, pickpocketing, bike thefts, thefts from cash machines, thefts from parking meters and diversion of electricity, while the ‘other crime’ category includes crimes such as fraud, handling stolen goods, failing to appear at court, dangerous driving, immigration offences and money laundering. The types of crime within each category are so different from other types of crime in the same category that any analysis of these categories as a whole would be meaningless and might lead to misleading results.

If the Home Office releases more detailed statistics in the future, this site will include more types of crime.

How are the maps constructed?

Creating hotspot maps for crime is difficult and it is very easy to produce misleading results, particularly if you are producing maps for the public to use. There is no one perfect system for identifying crime hotspots on maps, but this site uses a procedure that is currently accepted as being better than most. For each type of crime shown on the site:

  1. The approximate crime locations published on data download page of plotted on a map.
  2. A 250-metre grid is placed over the crime locations and the number of crimes in each grid square is counted.
  3. A kernel-density estimation (KDE) procedure (with a bandwidth of 500 metres) is used to identify hotspots from the counts of crimes in each square.
  4. The Getis–Ord Gi* test (calculated using a fixed-distance band with a threshold of 500 metres) is used to identify which crime hotspots are statistically significant, i.e. which hotspots contain more crime than would be expected in that area by chance.
  5. The Jenks natural breaks procedure is used to divide the hotspots into four groups according to the value produced by the KDE procedure. Each group is shown as a different shade on the final map.

The maps on this site are stored on GeoCommons, so you can download them or add them to your own website as long as you comply with the Creative Commons Attribution 3.0 Licence.

Why doesn't Crime in London work with my browser?

The maps and charts on this website use technologies that will not work in older web browsers. Providing separate fall-back options for out-of-date browsers would be time-consuming and expensive, and (since fewer and fewer people use older browsers) ultimately futile.

If you can, please upgrade your browser. If you are accessing this site from an institutional network and you are not allowed to update your browser, please activate Google Chrome Frame to improve your experience. In that case, please also contact your IT department and ask them to upgrade your browser. If you are not sure what browser you are using, find out.

Contact us

Use this form to contact me and I will reply as soon as we are able. All fields are required.