9th Crime Mapping conference, Pittsburgh
Details from the 9th Crime Mapping Research Conference, Pittsburgh, PA, 28-31 March 2007.
Index > conference notes > MAPS9
Jump down to the weighted displacement quotient calculation spreadsheet
Jump to the paper on near repeat shooting patterns in Philadelphia (Ratcliffe and Rengert)
Eric McCord, Travis Taniguchi and myself presented on, among other things, a revised approach to conducting a buffer analysis. It is perhaps best demonstrated with an example. In the street layout below, we are interested in measuring the intensity of crime events (red dots) around a potentially criminogenic location (green dot) up to a certain distance out from the location. The distance is termed the buffer distance. For example, the green dot could represent a bar, the red dots could represent late night assaults, and the buffer distance, shown here as a blue disc, could be a few hundred feet.
The traditional approach to a buffer analysis is to count every crime incident inside the buffer as having a value of one, and every event outside the buffer as having a value of zero. This is usually the case irrespective of the distance of the crime event from the criminogenic location. As shown below, there are a number of crimes within the buffer, but the three highlighted all count as a score of 1 to the calculation of the number of crime events within the buffer.
The tool we have developed calculates the intensity value in a different way. Instead of counting every crime as 1, each crime value is weighted so that crime events that are further from the center of the buffer count less than crimes close to the criminogenic site. The tool provides 5 different ways of measuring this inverse distance weighting. The example below shows a quartic kernel approach, but exponential and linear methods are also available.
Further instructions and the download page for the tool can be found here.
Weighted Displacement Quotient calculator spreadsheet
Clairissa Breen and I presented a paper titled "Targeted Law Enforcement in Camden, New Jersey: Diffusion of Benefits and Displacement of Crime". Given that displacement of crime can be a public relations nightmare for police departments, understanding and measure displacement or diffusion of benefits in essential. It seems inevitable that as soon as police target crime in one neighborhood, someone in another neighborhood will be seen on the evening news complaining that all the police are doing is moving crime from that neighborhood into our neighborhood. The weighted displacement quotient compares how much crime occurs between the targeted area (A), the buffer area (B) and the control area (C) before and during the operation to see how crime in those areas has changed. If crime has been displaced into the surrounding buffer or if that surrounding buffer has actually experienced some benefits due to its proximity to the operation, the WDQ will be able to detect that change. For further information on the WDQ, see: Bowers, K.J. and Johnson , S.D. (2003) 'Measuring the geographical displacement and diffusion of benefit effects of crime prevention activity', Journal of Quantitative Criminology, 19:3, pp. 275-301.
The Excel spreadsheet available to download from this site, calculates the weighted displacement quotient as well as a number of other calculations. All you have to do is feed in the amount of crime in three areas before and during (or after) a police operation. The three areas are a target area, a buffer area, and a control area.
On the first page (shown here) you enter values into the yellow boxes. Stage one asks for crime counts for the target area, a buffer area, and a control area both before and during (or after) the operation. At stage two, you can enter the name of the operation, or if you prefer, the name of the police district or area.
Once you can done that, click the yellow results tab shown at the bottom of the screen. This will take you to the results page where you can print the output and see the result. An example is shown below.
Some things about the spreadsheet:
The SEPTIC program is designed for reasonable numbers. Cells filled with a value of zero will result in a warning on the results page. The statistics used in this program will not function with a zero value in the control area.
Proceed to the download page
Near repeats in Philadelphia shootings
The paper that the Ratcliffe and Rengert presentation refered to is available for download in draft form from this location. Please select the Ratcliffe and Rengert paper when you get there.