GENERAL THEORY OF SPATIAL CRIME PATTERNS*

• DOI: http://doi.org/10.1111/1745-9125.12117
• Date: 01 November 2016
• Author: JOHN R. HIPP
• Published date: 01 November 2016

GENERAL THEORY OF SPATIAL CRIME PATTERNS∗

JOHN R. HIPP

Department of Criminology, Law and Society and Department of Sociology,

University of California, Irvine

KEYWORDS: neighborhoods, spatial, theory, ecological

I propose a general theory for examining the spatial distribution of crime by specif-

ically addressing and estimating the spatial distribution of the residences of offend-

ers, targets, guardians, and their respective expected movement patterns across space

and time. The model combines information on the locations of persons, typical spatial

movement patterns, and situational characteristics of locations to create estimates of

crime potential at various locations at various points in time and makes four key con-

tributions. First, the equations make the ideas involved in the theory explicit, and they

highlight points at which our current state of empirical evidence is lacking. Second, by

creating measures of spatial “potentials” of offenders, targets, and guardians, this the-

ory provides a precise grounding for operationalizing spatial effects in studies of place

and crime. Third, the equations provide an explicit consideration of offenders and

where they might travel and, therefore, incorporates offenders into crime-and-place

research. Fourth, these equations suggest ways that researchers could use simulations

to predict stable patterns, as well as changes, in the levels of crime at both micro and

macro scales. Finally, I provide an empirical demonstration of the added explanatory

power provided by the theory to a study of place and crime.

A burgeoning field of research has been focusing on the spatial distribution of crime.

One large research area has emanated from the Chicago School and has focused on

the ecological distribution of crime in meso units such as neighborhoods (Browning

et al., 2010; Hipp, 2007; Peterson and Krivo, 2010; Sampson and Groves, 1989; Sampson,

Raudenbush, and Earls, 1997). This tradition has often explored these spatial patterns of

crime through the lens of social disorganization theory. A second large research area has

grown in more recent years and has focused on crime at micro-locations such as street

segments or at even smaller units such as parcels (Weisburd et al., 2004; Weisburd, Groff,

and Yang, 2012; Wikstr ¨

om et al., 2010), which most frequently has employed routine

activities theory or crime pattern theory in studying these micro-units. Some researchers

have suggested ways to integrate both routine activities and social disorganization

theories (Smith, Frazee, and Davison, 2000).

In this article, I propose a spatially explicit theory that adopts insights from these

existing ecological theories and formalizes these various insights mathematically. So

although elements of the theory presented here are abstract, there are several advantages

to presenting these ideas in a formalized manner that mathematically defines these

propositions. First, these equations make the ideas involved in the theory unequivocal.

∗Direct correspondence to John R. Hipp, Department of Criminology, Law and Society, University

of California, Irvine, 3311 Social Ecology II, Irvine, CA 92697 (e-mail: john.hipp@UCI.edu).

C2016 American Society of Criminology doi: 10.1111/1745-9125.12117

CRIMINOLOGY Volume 54 Number 4 653–679 2016 653

654 HIPP

Second, these equations will clarify points at which our current state of empirical evidence

is lacking. In what follows, there are certain equations in which we have little evidence

regarding what the parameter values might be and, therefore, highlight areas of needed

research. Third, by creating measures of spatial “potentials” of offenders, targets, and

guardians, this theory provides a precise grounding for operationalizing spatial effects in

studies of place and crime. This approach moves beyond the unit of analysis problem and

moves beyond treating spatial effects simply as a nuisance. Fourth, the equations provide

an explicit consideration of offenders and where they might travel. Existing ecological

studies of crime typically have ignored the presence of offenders, and this theory provides

estimates of where offenders might be, which allows for empirical tests of levels of crime

at micro and macro scales as a consequence. I provide an empirical demonstration

of the added explanatory power to a study of place and crime. Fifth, these equations

suggest ways that researchers could use simulations to predict stable patterns, as well as

changes, in the levels of crime at both micro and macro scales. By using these equations

within a simulation framework, various predictions can be made and subsequently tested

empirically (Birks, Townsley, and Stewart, 2012; Groff and Mazerolle, 2008).

The proposed general theory of spatial crime patterns focuses specifically on the

spatial processes of crime. It takes these processes into account by formalizing them in

a series of propositions and then draws out implications. This model uses the frame-

work of routine activities theory (Cohen and Felson, 1979) to focus on three types of

persons moving about in the environment: potential offenders, targets, and guardians.

Routine activities theory posits that crime events will only occur when there is a co-

occurrence in time and space of a motivated offender, a suitable target, and the lack

of a capable guardian. As a result, a need exists for both offenders and targets to come

together in space and time. What is required, then, is information on where offenders,

targets, and guardians live and on where they go during the day and night, which

provides the starting point for this theory, and then these ideas can be formalized in

equations.

From one point of view, we would collect information on where offenders, targets, and

guardians are located at all times of the day. With this information, we could estimate the

equations that predict the amount of crime in a location based on these three key ingredi-

ents. Nevertheless, whereas existing research has attempted to proxy for the presence of

offenders, the presence of targets, the presence of guardians, or even sometimes the pres-

ence of two of these three ingredients, research has rarely tried to estimate the presence

of all three elements simultaneously. One example of a study attempting to measure the

movement of targets and guardians used the locations of schools and workplaces to model

the presence of students or workers at block locations during daytime hours and census

information to capture the nighttime presence of persons (Boessen, 2014). Furthermore,

with advances in various technologies such as global positioning system (GPS)-enabled

smartphones or cameras to record the presence of persons, it may become more feasible

over time to measure the presence of persons at various times of day (although the

challenge remains of distinguishing among offenders, guardians, and targets).

Rather than attempting to measure the locations of persons, the focus of this general

theory of spatial crime patterns is to build a model explaining where people might

potentially be at any given time. These potentials are probabilistic, allowing researchers

to draw general inferences about where and how much crime might occur across spatial

locations. In this model, it is important to track the movements of different types of