Using GIS to Fight Crime

By David Shillingford and Jon D. Groussman

In the last decade, GIS has become a powerful crime prevention and investigation tool for mapping and analyzing crime patterns.

Maps have been part of the detective’s repertoire since the early 1900s. Yet until the 1980s, detectives were still analyzing crime patterns by positioning different colored pins on wall maps. Many crimes have been solved by representing and recognizing spatial patterns of crimes. But map boards had some major limitations.

Once the map was removed, the spatial analysis was lost unless photographed. Over time, the relationships of events were difficult to represent. Also, the map could easily become crowded. Data sets could not be added or removed, and scale could not easily be changed. Those limitations and others were overcome as new technologies were applied to crime mapping.

The introduction of the mainframe computer in the 1970s allowed large amounts of data to be processed and software to be developed that visually represented maps and the location of criminal activity. However, high costs and limited access to mainframe computers made crime mapping a somewhat restricted and specialized practice until personal computers became widespread and affordable. In the 1990s, the popularity of the personal computer — allied with the commercial development of mapping and analysis software — caused crime mapping to become widespread. The 1994 Violent Crime Control and Law Enforcement Act also provided a boost to the implementation of GIS by providing funding for crime prevention programs. More recently, businesses have begun developing crime data analytics to understand and mitigate crime risks.

Figure 1:
Mapping Crime Risk

CAP Index CRIMECAST reports provide a quick and thorough overview of the crime risk at any given address. A detailed map and spreadsheet of risk scores isolate and identify crime-related issues around a specific site (in this example, 545 Washington Blvd., Jersey City, N.J.).

Current Scores (2009) National State County CAP Index 331 306 145 Homicide 319 316 190 Rape 83 148 173 Robbery 384 320 141 Aggravated assault 402 379 161 Crimes against persons 382 357 158 Burglary 48 114 130 Larceny 836 1248 697 Motor vehicle theft 123 150 89 Crimes against property 638 1046 613 CAP Index National State County Past – 2000 246 208 105 Present – 2009 331 306 145 Projected – 2014 319 322 155 CRIMECAST scores range from 0 to 2000 and indicate the risk of crime at a site compared to an average of 100. A score of 400 means that the risk is four times the average, and a score of 50 means the risk is half the average.

Click to enlarge

Crime Mapping Today
Crime mapping compares crime data in time and space against other human and environmental factors and events to establish patterns, develop leads, and predict future criminal trends. Data that can enhance crime mapping includes:

• different sets of crime data, such as comparing drug seizures and vehicle crimes (vertical correlation)
• home address of criminals and crime victims
• noncriminal environmental and human data (horizontal ­correlation), such as:
  • locations of common crime scenes (industrial complexes, convenience stores, or gas stations)
  • demographic data (age distribution in an area)
  • properties that might correlate to criminal activity (those with a liquor license)
  • topographic data (line of sight to a crime scene to canvass for witnesses)
  • economic data (home foreclosure rates that might correlate to arson cases)
  • time variables (season or day of the week)

Tools that enable and enhance the spatial analysis of crime data have fundamentally altered modern policing and are a large part of the drive toward intelligence-led policing. GIS can enhance public safety operations through:

• strategic planning
• improving the efficiency of resource allocation and ­accountability
• facilitating community policing
• presenting crime data in a compelling manner as part of a prosecution
• encouraging cross-jurisdictional intelligence sharing and cooperation

The Internet and Community Policing
The Internet has made crime mapping software more widely available and also provides the facility for law enforcement to share crime data with the community through online interactive maps. Much debate surrounds the pros and cons of greater transparency of crime data, particularly where data is incorrect and causes reputational damage or financial loss. Despite concerns over privacy, the trend is likely to grow as more companies offer the service to police departments — and given the public nature of much crime data.

Crime Mapping in Business Loss Prevention
Businesses are increasingly applying crime mapping to understand and mitigate criminal risks to their property and employees. Police departments typically address all crimes that fall within their jurisdictional boundaries. However, businesses tend to be concerned with certain types of crime over a wider geographical area — particularly businesses that have multiple sites in more than one police jurisdiction.

This presents a challenge for commercial users of crime data, as definitions and practices between different jurisdictions often vary, making it difficult for companies to compare like with like. For example, virtually every act of larceny may be reported to and recorded by the police in one jurisdiction, but only a fraction of such crimes are reported and recorded in another. Or the police might make crime data available to the public in one jurisdiction and not in another. Those variations may also affect a single-location business that is on or near a jurisdictional boundary. The solution is to apply statistical techniques to model crime and loss risk that underlies the varying crime and demographic patterns across the nation, similar to the CAP Index system (see Figure 1).

User Beware
Any technology, particularly an emerging one, has its limitations. To use GIS properly in criminal analysis, it’s critical to have a clear understanding of those limitations. Like any statistical model, the output is only as good as the data that goes into it. However, an additional note of caution applies to GIS modeling because people tend to be more persuaded by visual representations of information than when viewing numbers alone. Often with GIS models, the wrong scale is chosen. As a result, data will by concentrated and may appear to indicate that a crime hot spot exists when it does not. It’s also easy to put too much data into a single map, thereby making it confusing.

The Future
The next five years will see a convergence of technologies related to GIS — Global Positioning Systems (GPS), advanced analytics, predictive modeling, and the Internet — that will allow crime data to be analyzed in near real time and will help predict future activities of criminals. There will be significant growth in the public’s access to local and national crime data maps that will impact GIS in crime mapping. And public/private partnerships will be enhanced by the development of industry-specific information-sharing systems to link theft victims in crime verticals to law enforcement specializing in the investigation of such crimes. Those verticals include mobile equipment theft, cargo theft, and organized retail crime.

David Shillingford is president of ISO Crime Analytics Inc., an ISO company that helps insurers and their policyholders predict, mitigate, and recover from a variety of crime-related risks. Within ISO Crime Analytics, the National Equipment Register helps equipment owners combat construction and farm equipment theft, CargoNetTM enhances supply-chain security, and Enabl-u provides risk management services to the retail and banking industries.

Jon D. Groussman, J.D., is president and COO of CAP Index, Inc., a company responsible for identifying, forecasting, and mitigating crime and risk to 81 of the Fortune 100 companies. He is nationally recognized for advising private corporations and government entities on crime prevention strategies, as well as post-incident response and analysis.