Neighbourhood crime map: methodology and limitations

A brief description of how the neighbourhood crime map data was obtained and some of the limitations of the data.

The statistics used to create the crime map were provided to CBC News by the office of Toronto Coun. Denzil Minnan-Wong. His office had obtained the data, which was collected by Toronto police, from the city.

The names and boundaries of the 140 neighbourhoods depicted in the map are based on the City of Toronto's official neighbourhoods. Clicking on one of the neighbourhoods will reveal the statistics going back to 2004 for each crime category.

The colour gradient on the map is based on the range of criminal occurrences per 10,000 neighbourhood residents in each category in 2011.

For instance, there were 369 instances of assault reported in the Kensington-Chinatown neighbourhood in 2011. But the figure that was plotted on the map for that neighbourhood is 199.3, which is the number of assaults reported per 10,000 residents of Kensington-Chinatown in 2011, as reported in the census.

Per capita numbers are not displayed for past years because there is no reliable population estimate for those neighbourhoods for any of those years except 2006, which was another census year.

The "ranking" of a neighbourhood as it relates to criminal occurrences was also based on this per capita figure, and not the actual number of criminal occurrences in that area.

Limitations

  • There is a significant portion of criminal occurrences that have no associated geographical information. Those incidents have been laid out in a table below, as have the crime numbers for the city as a whole. Note that the city numbers include the occurrences that have no associated geographical information.
  • When it comes to drug charges in particular, there is a high proportion of ungeocoded occurrences. Drug charges with no associated geographic information accounted for 35.5 per cent of all drug charges in Toronto for 2011. "This could happen because the event occurred at a new address that has yet to be uploaded onto data tables or it was inputted as a general address, such as 'Eaton Centre' or ‘Near Yonge and Dundas,’" a police spokeswoman said in an email.
  • The murder section of the crime map should be interpreted cautiously, considering there were only 45 homicides in the city in 2011. With this relatively small sample size, it is not advisable, for instance, to conclude on the basis of murder data, that a given neighbourhood is significantly more prone to murders than another. Moreover, there were no murders at all in 100 of the 140 neighbourhoods.
  • Some crime indicators, notably, assault, encompass a wide range of criminal behaviour. There is no way of being able to judge the severity of assaults from the data that was made available to us. For example, there is no way of telling if a given neighbourhood has a high proportion of extremely violent assaults relative to the total number of assaults.

Crime indicators for Toronto

 20042005200620072008200920102011
Assault2444925420251782474323677230462286122759
Break and enter1735715172156181435412987124051192110961
Drug charges6340599272447029716566357002

7965

Murder6480718670626350
Robbery53495652591657105478544852185341
Sexual assault26292901270825372566268928512991
Stolen vehicle11102104571030798067869654255444385
Theft over $5,0001633135211881124104010381008989
Total6892367026682306538960852578655646855441

Crime indicators with no associated geographic information

 20042005200620072008200920102011
Assault19292146201918271179167317011672
Break and enter629498501481381463394426
Drug charges1561136718061707161717882772

2830

Murder45036111
Robbery689810690636523576530519
Sexual assault395558535495425450595524
Stolen vehicle124511741051718479385355220
Theft over $5,0001671601369177616782
Total66196718673859585287539764156274