When Crime Appears

The Role of Emergence

Jean McGloin, Christopher Sullivan, Leslie Kennedy, Jean Marie McGloin, Christopher J. Sullivan, Leslie W. Kennedy, Jean McGloin, Christopher Sullivan, Leslie Kennedy, Jean Marie McGloin, Christopher J. Sullivan, Leslie W. Kennedy

• Publisher: Routledge
• Year: 2011
• eBook ISBN: 9781136624483
• Topic: Social Sciences
• Subtopic: Criminology
• Index: Social Sciences

In recent years, the idea of emergence, which suggests that observed patterns in behavior and events are not fully reductive and stem from complex lower-level interactions, has begun to take hold in the social sciences. Criminologists have started to use this framework to improve our general understanding of the etiology of crime and criminal behavior. When Crime Appears: The Role of Emergence is concerned with our ability to make sense of the complex underpinnings of the end-stage patterns and events that we see in studying crime and offers an early narrative on the concept of emergence as it pertains to criminological research. Collectively, the chapters in this volume provide a sense of why the emergence framework could be useful, outlines its core conceptual properties, provides some examples of its potential application, and presents some discussion of methodological and analytic issues related to its adoption.

PART I

Introduction

CHAPTER 1

Moving Past the Person or the Context

Thinking about Crime as an Emergent Phenomenon

CHRISTOPHER J. SULLIVAN, JEAN MARIE McGLOIN AND LESLIE W. KENNEDY

The basic goal of criminological inquiry is to understand, with a high degree of confidence, why people offend and why crimes occur under some conditions but not others. From an objective standpoint, we often fall short of this goal. Though our theories and models allow us to identify some very consistent and robust predictors of crime, an honest assessment suggests that there is room for progress. Weisburd and Piquero (2008) recently reviewed a large cross-section of empirical work published in the discipline’s flagship journal, Criminology, to determine how well statistical models based on criminological theory actually explain crime. By using R-squared as an analytic benchmark, they found that the typical level of variance explained was modest. Equally as troubling, they found that our ability to explain crime and offending patterns has not improved over time, suggesting a general lack of scientific progress. This pattern framed a discussion about the degree of “uncertainty” in explaining crime and the curious void of attention given to that facet of the analytic process.

The ability to understand and predict social behavior perfectly is, of course, unrealistic, but that is hardly an excuse for lackluster progress. The room for improvement is clear and without debate: Weisburd and Piquero (2008) concluded that criminological studies typically leave between 80 and 90 percent of the variance in crime unexplained. In most sciences, explanatory power consistently this low would be a marker of problems in its theoretical foundations. At the same time, the disconnect between data and theory may reflect problems in how scholars conceive of and study the outcome of interest. On this note, Wikström (2008) suggests that these shortcomings in explaining crime result in part because the discipline has lost sight of the search for causal mechanisms in favor of empirically testing (and retesting) the correlates of crime. Such an approach “eases methodological and conceptual burdens” (Hughes, 2006: 44), but it also moves the field away from the importance of context and process, which fundamentally informed many of the earlier, seminal explanations of crime and offending (Abbott, 1997; Short, 1998). In short, as Wikström has noted, statistical covariates of offending, no matter how robust or consistent, are not necessarily generative factors for crime. He argues that we will be better off once we have a clearly defined conceptualization of crime to explain and begin to understand the processes and interactions that produce that outcome. In other words, perhaps it is time to revisit the way in which we frame our conceptions and studies of crime.

This issue is the core motivation for this book. We propose that if criminologists want to understand crime more completely, they should consider adopting a different, and relatively novel, conceptual and empirical framework. Both theory and research often segregate the “ingredients” of crime into separate units for discussion and inquiry, typically the person or the situation (Gottfredson, 2005). In an obvious way, this reduces our ability to understand crime because only one of its parts is under consideration when we know that both are necessary. In a less obvious way, this accepted practice can also limit progress because studying the components of crime in empirical isolation carries the assumption that the whole—that is, the criminal event—is based on the sum of these components (Holland, 1998). To be sure, individuals and situations are linked in an additive way through routines and lifestyles. For instance, substance use, spending leisure time in unstructured settings, and sharing social spaces with other crime-prone individuals have all been linked to a generalized propensity for impulsivity and deviance (Gottfredson and Hirschi, 1990), and are also known to introduce individuals into situations with more opportunities for delinquency and crime (Felson, 2002; McGloin et al., 2007; Osgood et al., 1996). However, scholars have also acknowledged for some time that crime emerges from an interaction of many components—that is, the elements that serve as precursors to crime are also related to each other in complex, nonlinear ways.

Kennedy and Van Brunschot (2009) argue that to comprehend the criminal event and/or patterns of crime properly, one has to articulate the elements that comprise the threat of its emergence, think carefully about the process whereby the event occurs or does not occur (as well as the nature of the event), and recognize that there are feedback effects of this event (or non-event) on the precursors that may affect later events. They advocate using a risk framework that takes a broad and inclusive view of the elements leading to and resulting from crime (thereby shedding light on crime patterns, not just singular events). This provides some definitional advantage as “risk” is a boundary-spanning concept.

The confusion that derives from a myriad of explanations and prescriptions about crime can be reduced somewhat when we draw in the common elements of risk … Risk … provides a metric … that can help us tie different parts of the crime problem together. It offers a probabilistic interpretation to crime analysis that allows us to suggest that certain things are likely to happen.

(Kennedy and Van Brunschot, 2009: 11)

Regardless of their theoretical preferences, most crime scholars would likely be comfortable discussing:

1 an individual’s risk for engaging in crime (and what affects that variation, such as low-self control) in given circumstances and over time;

2 situational risk for crime (along with what affects it, such as the presence of suitable targets or potential co-offenders);

3 how the interaction of the two affects the risk of criminal events as well as giving rise to crime patterns; and

4 how the outcome(s) of this interaction can have feedback effects on both the individual and situational risk.

The concept of emergence helps us understand how we can conceive of and study the production of criminal events and broader patterns of crime at places or across careers by fusing these elements of risk. Furthermore, the emergence perspective can frame explanations about single criminal events, or larger patterns and trends. Importantly, it also illuminates some of the struggles in understanding crime to this point. Viewing crime as an emergent phenomenon takes us away from linear explanations of events or behavior, focusing instead on complex interactions among the key elements that give rise to an outcome. In understanding crime there are “myriad interdependencies” that are not always simple in their manifestation (Felson, 2006: 61). As Holland (1998: 121–122) states, emergence is “a product of coupled, context-dependent interactions” where “the behavior of the overall system cannot be obtained by summing the behavior of the constituent parts … However, we can reduce the behavior of the whole to the lawful behavior of its parts, if we take the nonlinear interactions into account.” The emergence framework assumes that properties observed at a higher level are a manifestation of the interactions of lower-level components (Sawyer, 2001).

The origins of the emergence perspective are often traced to the work of John Stuart Mill during the 1840s (McLaughlin, 1992). Mill specifically focused on “heteropathic laws,” where the notion of an additive “composition of causes”did not appear to hold. Instead, examples of such laws were evident in situations where what was observed could not be fully deduced to its components (e.g., water) (Mill, 1868). Consequently, simple rules can be used to understand emergence based on interactions among lower-level elements, but emergent properties lack full reducibility (McLaughlin, 1992), which ensures some degree of unpredictability. Still, as the concept and its use have evolved, it is now recognized that there is a degree of predictability that can be attained in defining emergent properties through the process of empirical research and theoretical development (Hempel and Oppenheim, 1965). In drawing on the emergence framework, there is also a focus on interdependency and interaction within context (Wimsatt, 1997) as well as a clear structuring across different levels of understanding (Kim, 1999) and focus on mechanisms as they fit with theory and research (Crutchfield, 1994). In sum, with emergence, there is a sense that something new has arisen or appeared at another level based on the complex interaction of elements, forming a pattern where a certain degree of organizational cohesion can be clearly identified.

The emergence concept is also not totally new to criminology (although it has not always been specifically identified as this in the research literature) as its application has influenced crime theory for generations, beginning with the human ecologists. Because ecologists suggested rules for how social interactions would take place in urban areas, Abbott (1997) argued that their major task was describing interactional fields. Renewed interest in the Chicago perspective in part derives from a concern with interdependence, which reflects this interaction and the richness it brings to our understanding of social behavior. As Abbott (1997: 1152) puts it, “The Chicago school thought that no social fact makes any sense abstracted from its context in social (and often geographic) space and social time … Every social fact is situated, surrounded by other contextual facts and brought into being by a process relating it to past contexts.” He is critical of contemporary efforts to search for individual-based variables that can be analyzed in terms of independent–dependent relationships, largely because such exercises are devoid of concerns about the social contexts in which these behaviors take place. In large measure this would sum up the scope of the studies reviewed by Weisburd and Piquero (2008) and critiqued by Wikström (2008). The emergence perspective asserts that individual actions are heavily conditioned and interactive, making it difficult to explain or predict heterogeneous behavior in response to similar elements based on a series of additive estimates identified and averaged across a full sample.

One of the key tasks in an emergence framework is determining the rules of interaction—under what conditions will crime occur and under what conditions will it not occur (Corning, 2002; Holland, 1998)? This notion of simple rules that, in concert, generate a higher-order outcome is the core objective of understanding emergence (Holland, 1998). Focusing on process and interactions necessitates the articulation of how individual and situational risks come together. It is not enough simply to say they both matter. Instead, one must offer guidelines for the ways in which these planes of risk relate to each other and how they effectuate the mechanisms by which crime occurs. While the notion of emergence calls to mind complex systems, and crime certainly qualifies as complex, we can still identify relatively straightforward characteristics of individual action that give rise to these outcomes.

In framing crime as a situated transaction, Wilkinson and Fagan (1996) stress the importance of integrating information about individuals in the setting, characteristics of the settings, and the rules and contingencies that dictate their interaction (see also Decker, 1996; Horney, 2006; Luckenbill and Doyle, 1989; Kennedy and Forde, 1999; Short, 1998). A similar view was recently articulated by Wikström (2006: 87), who argued that “individuals’ knowledge and skills … are always applied to the particularities in the setting in which they take part. It is the interaction between an individual and his environment that determines his course of actions (inactions).” In other words, much like many other phenomena, crime is somehow different from and more than the sum of its parts. Embracing this notion, rather than singularly considering what constitutes individual risk or situational risk, could be a fruitful avenue in increasing our ability to understand crime.

And, though the emergence perspective recognizes that the rules do not always generate the same outcome, engendering some uncertainty, we can identify regularities that help in explaining the outcome of interest. This does not necessitate “new” theory, but instead can guide (1) more precise articulation about the assumptions of extant theory and (2) the exact ways in which particular theories should be integrated. This enhanced understanding of causal interactions provides a means of better identifying the processes that lead to crime (Wikström, 2008). So, to some extent, adopting an emergence perspective harkens back to an earlier age of criminology, when the focus hinged on causal processes and mechanisms, with an eye toward complex social interaction and varied levels of explanation (Cloward and Ohlin, 1960; Shaw, 1931; Shaw and McKay, 1942; Short and Strodtbeck, 1965; see also Short, 1985).

When thinking about crime in an emergence framework, the building blocks of “individual” and “situational” risk must first be defined. This process will largely be informed by one’s theoretical preference—in other words, the emergence framework is applicable for peopl...

Table of contents

• Cover
• Title
• Copyright
• Contents
• Series Editors’ Foreword
• Preface and Acknowledgments
• PART I. Introduction
• PART II. Explaining Crime
• PART III. Crime Emergence in Action
• PART IV. Studying Crime Emergence
• Notes on Contributors
• Index