As psychological profiling seems to garner all the attention with television shows that feature FBI profilers, one should be reminded that there is another method of criminal profiling that, while less glamorous, seems to provide a very high degree of reliability. The ‘other’ method is geographic profiling that had it beginnings in a London crime spree involving a serial rapist. In these cases occurring in the mid-1990’s, stumped police called in David Canter, Professor of Psychology at the University of Surrey to assist. After analyzing crime data superimposed over a map, Professor Canter made the suggestion that the perpetrator was to be found in a very specific geographic area based upon these furloughs of crime (Wilson 2002, pp. 14-16).
Subsequent retroactive studies have confirmed a number of precepts regarding the proposition that one can predict the “base” of criminals from computerized algorithms that use generally one of two validated heuristic methods. In fact, of the area defined by a “search” algorithm that states that is inclusive of a circle defined by the two furthermost crime scenes, fully 51% of the serial rapist offenders lived within the first 5% of the defined area and an overwhelming lived within 87% of the first 25% of the defined search area (Canter, Coffey, Huntley, & Missen 2000, p. 457). This algorithm holds regardless of the crime as indicated by the table below (Canter, et al 2000, p. 458):
% Living in Circle
In addition to the “circle” algorithm described above, another method utilized is the “Decay” method with analyzes data from the assumption that perpetrators commit crimes near where they live. This method is based upon the idea that crimes ‘decay’ as the distance to the criminal’s residence decrease, in effect, creating something of a ‘buffer zone’ immediately surrounding their house (Snook, Taylor, & Bennell 2004, p. 107; Canter, Coffey, Huntley, & Missen 2000, pp. 459-460).
Somewhat against these two algorithms are the approximately 10% of perpetrators that are labeled as “commuters” and do not live near the crime scene (Snook, Taylor, & Bennell 2004, p. 117). Nevertheless, in these cases as in the easier to classify “marauder”-type crimes, there is the idea that criminals operate from a base of familiarity such as a home or workplace or other personally significant locale or the route that connects them. While focusing on geography, this brings into the consideration the important psychological construct of “mental maps” in that the location must have significance from the perspective of the perpetrator… it need not be an actual geographic landmark to have psychological significance that plays out geo-centrically (Holmes & Holmes 1996, p. 152). As such it is no surprise that both methods are reliable and that there is a significant degree of overlap in the two model predictions.
Of greater interests to both researchers and law enforcement are “serious” crimes such as murder, rape, arson and burglary. While there is likely a logical high correlation between other ‘petty’ crimes such as speeding tickets or parking violations due to the correlates of where one most spends time is of course most likely a location where there is a high likelihood of violation. With regards to the applicability of the model with regards to even serious crimes, the model is based upon serial crimes, that is, a perpetrator who does so multiple times. Research suggests that as criminals gain “experience” there is a statistically significant correlation with the distance from home the criminal is (Snook 2004, p. 53). Further, in regards to burglary, there is the additional significant correlation of the value of the crime with the distance traveled to commit the crime (Snook 2004, p. 55). While difficult to put a value on personal crimes, this finding may extend to crimes such as murder, arson or rape in which it is conceivable that victims could have some relative value to the criminal, similar to the monetary value gained from burglary. Such a concept may partially explain that, as these types of criminal gain experience, they tend to venture out further from their base as they commit additional predatory crimes. In terms of geographical profiling and the case in which an investigator believes that a number of crimes have been committed by a single person, this has special significance as it can serve to focus the hunt for the offender in the vicinity of the first few crimes (Holmes & Holmes 1996, p. 155).
Of particular interest is that these methods can be utilized without the computer or complex mathematical algorithms. Studies indicated that, in lieu of years or experience or ‘gut feelings’, incorporating these methods simply as heuristic models whose basic precepts can be taught in a class, students can nearly equal the computerized version (Snook, Canter, & Bennel 2002, pp. 116-117).
Though all law enforcement operates with the goal of catching the perpetrator, authorities do not have unlimited resources of time or finances to investigate all leads thoroughly. Geographical profiling also serves to focus the attention and other resources of an investigative department into the area in which validated statistics indicate there is is a pay-off. By using these methods, search costs can be minimized and offender can be apprehended faster, ideally prior to the accumulation of an additional victim that shows up as a simply data point on a geographic profile (Canter, Coffey, Huntley, & Missen 2000, pp. 459-460).
Canter, D., T. Coffey, M. Huntley, & C. Missen. (2000). “Predicting Serial Killers’ Home Base Using a Decision Support System”. Journal of Quantitative Criminology, (16), 4, pp. 457-478.
Holmes, R. & S. Holmes. (1996). Profiling Violent Crimes: An Investigative Tool, 2nd Edition. Sage Publications: Thousand Oaks, California.
Snook, B., P. Taylor, & C. Bennell. (2004). “Geographic Profiling: The Fast, Frugal and Accurate Way”. Applied Cognitive Psychology, (18), pp. 105-121.
Snook, B., D. Canter, & C. Bennel. (2002). Predicting the Home Location of Serial Offenders: A Preliminary Comparison of the Accuracy of Human Judges with a Geographic Profiling System”. Behavioral Sciences and the Law, (20), pp. 109-118.
Snook, B. (2004). “Individual Differences in Distance Traveled by Serial Burglars”. Journal of Investigative Psychology and Offender Profiling, (1), pp. 53-66.
Wilson, J. (2002, November). “Mapping Murder”. Geographical, (74), pp. 14-17.