Nancy G. La Vigne
Crime Mapping Research Center
U.S Department of Justice

POSITION STATEMENT

What are your research interests in the areas of inequality and equity, and what spatial dimensions do you currently or potentially see in them?

From a criminological perspective, the issues of inequality and equity relate to the spatial dimensions of crime and criminal behavior and associated demographic and physical characteristics of the environment. Known patterns of the spatial distribution of crime and criminal behavior have been associated with three broad areas of inequality: (1) geographic concentrations of crimes affecting specific (disadvantaged) neighborhoods and populations over others; (2) social and economic inequalities resulting from the unequal distribution of criminal justice policies and services; and (3) the disparate impact on neighborhoods and communities due to both the incarceration of family members and the release of offenders back into these communities.

Spatial analyses examining the relationship between crime and socio-economic status date back to cartographers in France and England who, as early as 1830, were examining the influence of wealth and population density on crime (Quetelet, 1842). In the United States, social ecologists in the 1940s mapped juvenile delinquents and analyzed relationships between delinquency and various social conditions (Shaw and McKay, 1942). Today, countless research studies examining these relationships exist, although until recently, vary few were conducted with sound spatial-statistical methods (See answer to question 2 below).

A less studied but equally important application of spatial analysis to inequity issues associated with criminal justice is that of the distribution of criminal justice resources and services. Coined "spatial injustice" by Rengert (1989), this area of inquiry concerns a variety of ways in which criminal justice policy has a disparate impact on certain populations. Law enforcement, for example, might be less likely to respond to calls from places they perceive to be dangerous, leaving law-abiding residents in the area at higher risk of victimization. Likewise, police interventions targeted toward one location or "hot spot" may result in displacement to adjacent areas that are not targeted for crime reduction (see Rosenbaum, 1988). Sentencing decisions also can have a disparate impact on certain communities: a study of death penalty cases nationwide found that the death penalty was meted out disproportionately according to region, controlling for the nature of the homicide and the criminal history of the offender (Harries and Brunn, 1978).

Related to sentencing practices, the criminal justice system�s decisions to incarcerate and release offenders can also have a disproportionate impact on certain geographic areas. Clear and Rose (1999), for example, examined the impact that incarceration of offenders has on neighborhoods. They found that high levels of incarceration of residents from certain neighborhoods suppressed reductions in crime. Other research has identified a detrimental impact on children and families whose fathers are incarcerated. These families often suffer economically and emotionally from removal of a breadwinner and caregiver/father figure from their daily lives (Clear and Rose, 1999; Gabel, 1992). The flip side to this area of inquiry is the examination of the effect that the return of ex-offenders released from prison has on high-incarceration neighborhoods. A qualitative study of this question through interviews with residents suggests that high rates of offender reentry into an area can create economic and quality of life hardships for the community (Clear et al., 2000).

What kinds of spatial data, models, techniques, software, etc. do you use or have considered using in your research. Which of these works well for you? Where do you see problems and/or shortcomings?

Criminologists tend to be trained�in part�as traditional statisticians and therefore historically, criminological studies that had a spatial dimension were nonetheless conducted with an absence of knowledge or application of spatial statistics. This has changed over time, much due to the greater prominence that the role of place has had in criminological theory, as well as to the availability and dissemination of new methods and software.

On a very basic level, the advent of GIS in a user-friendly PC environment has enabled researchers to explore visualization techniques such as buffer, intersect, union, and measure. Buffering can be used to examine the potential geographic displacement of crime following an intervention. Intersect and union tools enable one to examine relationships that may indicate likely causal factors, guiding further exploration with more sophisticated spatial statistics. Measurement tools facilitate the examination of how offenders� journeys to crime differ by both crime type and characteristics of the physical environment (e.g., street networks, natural and man-made barriers, etc.). Spacestat�s Exploratory Spatial Data Analysis (ESDA) functions have taken such visualization methods several steps forward by linking descriptive analysis results (such as outliers in boxplots) to geographic locations on a map, making the spatial identification of, for example, convenience stores with unusually high numbers of police calls-for-service extremely straightforward.

Complementing these exploratory visualization techniques, criminologists have begun to conduct regression analyses that include the all-important consideration of the complicating effects of spatial autocorrelation. Again, the use of SpaceStat for this purpose has helped promote more (and better) criminological research employing the spatial regression approach.

In addition, CrimeStat, a spatial statistics software program released in 1998 (with version 1.1. released in 2000) and designed specifically for the analysis of crime incident locations, has promoted greater rigor in spatial analyses of crime. CrimeStat includes: statistics for describing the spatial distribution of crime incidents (e.g., mean center, standard deviational ellipse, and Moran's I); statistics for describing properties of distances between incidents (e.g., nearest neighbor analysis); both statistical and kernel density routines for conducting "hot spot" analyses (e.g., hierarchical nearest neighbor clustering, K-means clustering, local Moran statistics, and surface or contour estimates of the density of incidents); and a journey to crime module for analyzing serial offenders. Perhaps the greatest contribution of CrimeStat is this last module, which calibrates a routine for identifying a travel distance function and an estimation routine for modeling the likely location of the offender using either the calibration function or a mathematical model. The ability to predict statistically the location of a serial offender based on past crimes has been welcomed by criminal justice practitioners and researchers alike.

Can you point out any "best practice examples" of spatially oriented research in your field? Do you have any suggestions for Learning Resources CSISS might provide? Workshops we might offer?

While great strides have been made in the spatial analysis of crime and criminal behavior in recent years, the percentage of criminologists engaged in true spatial analysis remains relatively low. Best practices exist in the form of practitioner applications that have successfully reduced crime, identified suspects, and supported prosecutions (see

La Vigne and Wartell, 1998; 1999). Academic research in this area has begun to show great promise, with some "high end" applications combining GIS with neural networks, spatial econometrics, and the use of feature space analysis into space-time prediction (see Liu and Brown, 2000). These predictive modeling efforts represent the future of crime mapping and have the potential for bridging research and practice; predicting crime hot spots before they emerge (and thus focusing law enforcement efforts on prevention) can have a significant impact on crime.

In terms of training workshops, I recommend partnering with Dr. Ned Levine, creator of CrimeStat, to provide a series of training sessions on the mechanics of CrimeStat and its application to criminological research.

References

Clear, T.R. and Rose, D.R. (1999) "When neighbors go to jail: Impact on attitudes about formal and informal social control." National Institute of Justice Research in Brief July, 1999.

Clear, T.R., Rose, D.R. and J. A. Ryder (2000). "Coercive Mobility and the Community: The impact of removing and returning offenders." Paper presented at the Urban Institute�s Reentry Research Conference, October 13, 2000.

Gabel, S. (1992). "Children of incarcerated and criminal parents: Adjustment, behavior, and prognosis." Bulletin of American Academic Psychiatry Law, 20:33-45.

Harries, K.D. (1999). Mapping Crime: Principle and Practice. Washington: National Institute of Justice, U.S. Department of Justice.

Harries, K.D. and Brunn, S. (1978). The Geography of Laws and Justice. New York: Praeger.

Liu, H. and Brown, D.E. (1998). "A new point process transition density model for space-time event prediction." Proceedings of the IEEE International Conference on System, Man and Cybernetics, October.

La Vigne, N.G. and Wartell, J. (1988, 1999). Crime Mapping Case Studies: Successes in the Field, Volumes 1 and 2. Washington: Police Executive Research Form.

Quetelet, A. J. (1842). A treatise on man. Gainesville, FL: Scholar�s Facsimiles and Reprints (1969 ed.).

Rengert, G.F. (1989) "Spatial justice and criminal victimization." Justice Quarterly 4: 534-64.

Rosenbaum, D.P. (1988). "Community Crime Prevention: A preview and synthesis of the literature." Justice Quarterly 5:323-96.

Shaw, C.R. and McKay, H.D. (1942). Juvenile Delinquency and Urban Areas. Chicago: University of Chicago Press.