Pub. Date:
University of California Press
Risk Terrain Modeling: Crime Prediction and Risk Reduction

Risk Terrain Modeling: Crime Prediction and Risk Reduction

by Joel M. Caplan, Leslie W. Kennedy
Current price is , Original price is $49.95. You

Temporarily Out of Stock Online

Please check back later for updated availability.

Product Details

ISBN-13: 9780520282933
Publisher: University of California Press
Publication date: 06/28/2016
Edition description: New Edition
Pages: 192
Product dimensions: 7.00(w) x 9.90(h) x 0.40(d)

About the Author

Joel M. Caplan is Associate Professor at Rutgers University, School of Criminal Justice.

Leslie W. Kennedy is University Professor at Rutgers University, School of Criminal Justice, where he served as Dean from 1998–2007.

Read an Excerpt

Risk Terrain Modeling

Crime Prediction and Risk Reduction

By Joel M. Caplan, Leslie W. Kennedy


Copyright © 2016 The Regents of the University of California
All rights reserved.
ISBN: 978-0-520-95880-7




Criminal behavior is best understood as a social product that occurs in a patterned fashion, rarely fluctuating wildly from time to time or place to place. This observation was first made 170 years ago by Quetelet (1984). We believe that this enduring pattern occurs because the underlying factors that increase or decrease the risk of crime are not quick to change and exert fairly consistent effects on the appearance, distribution, and persistence of crime by attracting illegal behavior. However, although this pattern appears to be fairly regular, if not chronic, over time at the aggregate level, there are many factors that contribute on the micro level to the ever-changing landscape of crime incidents. Of interest to us here are how these factors may combine to encourage crime to start, how they affect the momentum of crime events over time, and how they can be manipulated to make crime stop.

The ideas that were developed and discussed by Quetelet and others about the origins and persistence of crime took on new urgency with the massive growth of American cities at the turn of the twentieth century, due to large waves of immigrants who began to flow into the United States and other Western countries. These migrants brought about changes in urban areas that caught the attention of researchers who were concerned about the negative impact that this rapid growth was having on communities. The consequences for urban planning, social reform, and economic transactions were transformative. Accompanying these changes were new concerns about crime and delinquency. In the heady days of urban research that ensued, Clifford Shaw and Henry McKay began to map urban areas and emphasized contextual factors related to delinquency. Shaw and McKay (1969) used this contextual mapping approach to document the areas in which crime had persisted over time.

Human ecologists (Park, McKenzie, & Burgess, 1925) talked about "natural areas," a term that appeared in studies of delinquency in Chicago in the early twentieth century. Natural areas, according to these researchers, were settings that had certain characteristics that led to predictable behavioral outcomes. Shaw and McKay reported through methodical observation that "natural areas" in Chicago appeared constant over time. They plotted delinquency incidents in Chicago over many decades and found that they concentrated in "transitional" zones. In addition, they reported that crime declined as one moved from the inner-city areas to the (outer) suburbs. A key observation from their research was that community characteristics and problems (for example, cultural conflict, gang behavior, conflict with families) stayed the same despite the changing attributes of the inhabitants (Hatt, 1946). As people came from and went into these areas, the social disorder and delinquency remained high, despite changes in the ethnic composition of inhabitants. Oddly, despite the importance of Shaw and McKay's finding that community characteristics matter for delinquency and its reduction and prevention, they overlooked it in their prescriptions for addressing the delinquency problems that interested them. They suggested instead that the behavior of people in these areas defined their qualities despite the physical characteristics that these areas exhibited. As Snodgrass points out:

To interpret their findings, Shaw and McKay relied most heavily upon the general concept of "social disorganization," the breakdown of social controls in the "communities" located in the transitional zone. The invasion by business and industries from the center of the city into the former residential areas created a wake of social disorganization in its advance which disturbed social cohesion and disrupted traditional conduct norms. Shaw and McKay explicitly and repeatedly mentioned industrial invasion as a primary source of communal disorganisation, although other sources, e.g. the influx of successive waves of highly mobile immigrant groups, were additional contributing factors, though not unrelated to business expansion. (Snodgrass, 1976, p. 9)

Their emphasis on social disorganization made sense to Shaw and McKay as social activists who believed that the causes of delinquency resided within the local traditions and cultural values of the inhabitants, even though, again, they were quick to point out that as different groups passed through these areas (particularly the zone of transition), the problems of delinquency and social disorder persisted. In other words, the factors that stayed consistent in these areas, that is, businesses and other physical features, were treated as tangential to the ways in which delinquency emerged and areas deteriorated.

As Snodgrass further points out:

A most striking aspect of Shaw and McKay's interpretation, then, is the absence of attempts to link business and industrial invasion with the causes of delinquency. The interpretation stayed at the communal level and turned inward to find the causes of delinquency in internal conditions and process within the socially disorganized area. Thus, their interpretation stopped abruptly at the point at which the relationship between industrial expansion and high delinquency areas could have gone beyond the depiction of the two as coincidentally adjacent to one another geographically. (p. 10)

In fact, Shaw and McKay did not see proximity to industry and commerce as causal but rather simply as an index of the areas where delinquency would be located. This failure to account for the effects of community characteristics, or environmental features, in attracting illegal behavior and spurring crime is surprising, given their huge effort in identifying spatial patterns of delinquency through mapping incidents, a project that went on for over 40 years.

Bursik (1988) points out that stability in ecological influence stayed constant before World War II in Chicago but changed thereafter, thus affording the opportunity to compare how these ecological factors influenced criminal behavior. In addition, generalizing the influence of environmental factors to the experience in other cities was hard to achieve and led to criticisms that Shaw and McKay's approach was not replicable (p. 526). But the observation that environmental factors can influence the nature of places is important and should not be lost in the disappointment concerning the inability to replicate Shaw and McKay's findings within Chicago over time or in another city in a predictable fashion. The external validity problem appears to originate not from the conceptualization of the importance of environment but from the limitations in the methodology used to measure its effects. It also derives from a fixation on the actors in crime rather than a consideration of them in the spatial contexts in which they operate, a divergence in approach since Shaw and McKay's time that has persisted in crime research until recently.

So, despite their reliance on maps and time series data to illustrate crime persistence, Shaw and McKay did not fully explicate how environment ties to crime emergence or outcome. There are conceptual and methodological reasons for this that we will explore below. It should be noted that Shaw and McKay's assumptions concerning the importance of the links between neighborhood characteristics and crime have been extensively studied using a social disorganization perspective that concentrates on the ways in which social control manifests itself in certain locations, typified by poverty and high levels of in-and-out migration. In particular, the work of Shevky and Bell (1955) examined the ways in which family status, socioeconomic characteristics, and ethnicity combined to influence behavioral outcomes using social area analysis. Social area analysis improved on the inflexibility of the idea of natural areas by combining community features through the way that they overlapped in different locations (Hatt, 1946; Heitgerd & Bursik, 1987; Janson, 1980). (See figure 1.) The areas of overlap were considered the locations in which crime problems would be greater.

Unfortunately for social area analysts, as was the case with the ecologists who preceded them, they were unable to move beyond macrolevel explanations for delinquency outcomes in spatial terms based on the underlying characteristics in the study area. Their multilevel approach was novel, however, and was adopted by urban planners such as McHarg (1995) to help depict the concentration of features in a landscape.

Recent work on social disorganization has focused on the ways in which areas suffering from social and physical disorders respond through collective efficacy — the pooling together of efforts to extract resources to battle problems faced in neighborhoods (Morenoff, Sampson, & Raudenbush, 2001). This work has provided an important stimulus for community planners to think more broadly about how community empowerment can be used to combat serious consequences of disorganization. Operating at the community or neighborhood level allows for a comprehensive assessment of local well-being and elicits steps that can be taken to address inequality and social upheaval. However, this research still relies on aggregate statistics and tends not to account for the physical environment as a major factor, at the microlevel, in bringing about criminogenic conditions that regularly attract illegal behavior.


Helping to overcome the limitations outlined above, improvements in data collection and advanced mapping technology have opened up the possibility of better microlevel analysis of places and crime. However, with advances in geospatial approaches, the ways that features of a landscape have been modeled in a geographic information system (GIS) are often contrary to how people experience and conceptualize their environments (Couclelis, 1992; Frank & Mark, 1991). Geographers suggest that regions, such as cities, are learned piecemeal rather than imagined whole by humans over time, an assertion that is grounded in views from psychology (Freundschuh & Egenhofer, 1997; Montello, 1993). So when assessing the likelihood or risk of crime occurring at conceivably any location throughout a city landscape, vector objects in a GIS (for example, points that are used to represent things such as bars, schools, or bus stops) are poor representations of criminogenic features on a map because they bear no particular relationship to the dynamic environments of which they are a part (Couclelis, 1992). "There are difficulties with this view of the world," explained Couclelis (1992, p. 66), "mainly that points, lines, and polygons that define vector objects do not have naturally occurring counterparts in the real world." They are approximations of environmental features, but without any theoretical or empirical link to their geographies (Freundschuh & Egenhofer, 1997).

Broad inattention to different spatial conceptualizations of criminogenic features by crime researchers has led to misrepresentations of these urban, suburban, and rural features in geographic information systems and resulting maps (Freundschuh & Egenhofer, 1997). The way people (for example, motivated offenders or suitable victims) conceptualize and operate in space is an important consideration for the mapping of the risk of crime throughout landscapes. Cartographically modeling these conceptualizations and the spatial influences of criminogenic features in a GIS in a way that reflects the actors' views is an important part of what Freundschuh and Egenhofer (1997, p. 363) describe as "Naïve Geography, a set of theories of how people intuitively or spontaneously conceptualize geographic space and time" (Egenhofer & Mark, 1995). It can yield more meaningful inferences about criminal behavior and actionable spatial intelligence for use by public safety professionals (Frank, 1993; Mark, 1993; Freundschuh & Egenhofer, 1997). Spatial risks for crime must be considered in terms of how the environment forms behavior.


We will use the concepts of "space," "place," and "area" (and variations thereof, for example, "spatial") deliberately throughout this book. So to clarify: "Space" is defined as a continuous expanse within which things exist and move. "Place" is a particular portion of space where defined activities or functions may occur. A place is the microlevel unit of analysis for risk terrain modeling (RTM). An "area" is a part of space defined as two or more contingent places.

Examining places rather than people for crime analysis does not remove the importance of the human factor. It simply shifts the focus away from personal characteristics to personal preferences. How individual persons select and use the environments that they occupy and the impact that this has on crime outcomes become the direct focus of the spatial risk perspective. This approach to crime analysis suggests a way of looking at behavioral outcomes less as deterministic and more as a function of a dynamic interaction among people that occurs at places. The attributes of places that we seek to identify are not constant, nor necessarily are the interactions set in place over time. However, the ways in which these factors combine can be studied to reveal consistent patterns of interaction that align with the view expressed by Brantingham and Brantingham (1981) in their development of crime pattern theory.

Risk provides a metric that can help tie different parts of a crime problem together and offers a probabilistic interpretation for crime analysis that allows us to suggest that certain things are likely to happen and others can be prevented according to our risk assessments (Kennedy & Van Brunschot, 2009, p. 11). Risk is based on a consideration of the probabilities of particular outcomes. When opportunity for crime is thought of in terms of "risk of crime," places can be evaluated in terms of varying degrees of criminogenic risk relative to certain nearby or faraway features of the environment (Cohen, Kluegel, & Land, 1981; Caplan, 2011). Again, this directs attention away from a fixation on only the offender or victim in responses to crime and permits considerations of characteristics of places as well.

In their simplest form, place-based interventions lead to strategies that direct police to particular areas to use the tools most directly available to them to solve problems, such as arrests or summonses targeted at people located there. But this approach is incomplete. Problem-oriented policing has offered important clues on how we can change situations to make them less conducive to crime (Mastrofski, Weisburd, & Braga, 2010). In this regard it is important to address the collective influence of certain spatial features as a principal approach to crime prevention. In arguing for improving how we study crime events, Braga and Clarke (2014) present a compelling justification for studies of places that focus on risks associated with certain types of environmental features. These features can create opportunities for crime, attract offenders, enable illegal behavior, and confound agents of social control in containing or suppressing their negative effects. But, at the same time, an understanding of an environmental feature's relative importance in creating risk of crime, as well as an understanding of how to target these features, can offer geographically focused strategies for crime prevention. The spatial risk perspective not only addresses the role that changing situational factors might have on a crime outcome, but also evaluates the overall effect of addressing the relative risks presented by features that have strong spatial influences on criminal behavior. With RTM, we can identify these features and their interaction with others in creating risky places. These places should be key targets for change and crime prevention.


Excerpted from Risk Terrain Modeling by Joel M. Caplan, Leslie W. Kennedy. Copyright © 2016 The Regents of the University of California. Excerpted by permission of UNIVERSITY OF CALIFORNIA PRESS.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

Table of Contents

List of Figures
List of Tables

1. Explaining the Contexts of Crime
2. Risk Terrain Modeling Methods
3. Crime Emergence, Persistence, and Exposure
4. Presence, Repeats, and Concentration: Exposures to Crime
5. The Theory of Risky Places
6. Event Contexts of Risky Places
7. Risk Management and RTM in ACTION
8. Risk Reduction



Customer Reviews

Most Helpful Customer Reviews

See All Customer Reviews