In the last three decades, researchers of the relationship between race and crime in North America have broadened their focus beyond differences between blacks and whites to include other groups: for example, studies of Hispanics and Asians are increasingly common. However, one group is continually underrepresented in studies of crime and, particularly, studies of violence: American Indians, or, as they are known in Canada, Aboriginals. In the preamble to her groundbreaking study of violence among American Indians, Bachman (1992) notes that prior to her work, multivariate studies of causes of Aboriginal violence were virtually nonexistent. Four years later, Nielsen and Silverman (1996: xii) found little had changed since Bachman's research. They remark that,

This oversight is surprising. Although Native peoples represent a small proportion of American and Canadian populations, they are consistently overrepresented in these nations' arrest and incarceration statistics. For example, Silverman (1996) estimates that in the period between 1987 and 1992, less than 1 percent of the U.S. population was American Indian; however, in these years the arrest rate for violent crime among American Indians was approximately 50 percent higher than for white Americans; for homicide, it was about 70 percent larger (see also Bachman, 1992; LaFree, 1995; Snyder-Joy, 1995).

To understand street life and its connections with crime, it is important to examine circumstances that lead youth from their families, as well as their lives on the street. A sense of these background and foreground experiences emerges from interviews we conducted with street youth. In these interviews, street youth describe their pasts and provide insights into experiences and events that lead to leaving home. These interviews also offer an introduction to the pleasures and problems of street life, including companionship and the availability of drugs and alcohol, as well as the difficulties of finding shelter, safety, food, and money. Many of these stories are disturbing, but they add a penetrating personal dimension to our understanding of the life histories of street youth and the daily situations they encounter.

In this and following chapters, we draw on nearly one hundred of the more than two hundred interviews we conducted with street youth in the third wave of our summer panel study. We present background details or experiences and a pseudonym to distinguish among the many young people with whom we spoke. This sociobiographical information is introduced to develop a composite picture of street youth that parallels the personal narratives often provided for smaller numbers of subjects in qualitative studies.

We provide a background to the interviews by drawing first on descriptive statistics from our initial study of 390 street youth in Toronto. This purposive sample was developed between the springs of 1987 and 1988 and involved surveying youth in shelters, drop-in centers, city parks, and on street corners in downtown Toronto.

This book paints a mostly grim picture of the daily lives of urban street youth. While most young people invest the largest part of their daily energies in the relatively benign worlds of school or work, homeless youth spend most of their time less profitably and more dangerously on the street and in parks, social assistance offices, shelters, and abandoned buildings. Together with friends acquired on the street, they spend a large part of their time looking for food, shelter, and money. Some actively seek and find work, but most remain unemployed, spending their time hanging out, panhandling, partying, and foraging in the shadow economy of the street.

Clearly, the outlook for most of these youth is bleak. Most will experience profound difficulties in making the transition from adolescent to adult roles. Recent structural changes in employment opportunities and the increasing domination of the service sector with its poor-paying and unstable jobs make this transition especially problematic for youth with limited human and social capital (Krahn, 1991; Revenga, 1992). Life course research confirms that the problems of troubled youth often anticipate difficulties in adulthood, in terms of crime, work, marriage, and other measures of well-being (e.g., Robins and Rutter, 1990). This body of research suggests the odds are poor that street youth will successfully traverse the important transitions necessary to move successfully from adolescence to normative adulthood.

Although research reported in this book was conducted in Canada, this work was undertaken against a backdrop of theories of crime developed in Europe and the United States, and of research on the homeless in developing countries of Central and South America as well as in North America and Western Europe. This chapter links the pieces of this international story together and addresses theoretical and research issues associated with the study of youth and crime. One of the most salient of these issues is the range of theoretical orientations we use in studying street youth. Some of these perspectives – for example, strain and control theory – usually are seen as incompatible.

We propose that strain, control, and other prominent theories can be bridged within a social capital theory of crime. Our perspective contributes to a growing body of work that suggests advantages of integrating theories of crime (Johnson, 1979; Elliott et al., 1985; Messner et al., 1989), but our approach is distinct in its use of an overarching concept to unite these diverse explanations. To set the context for our theoretical synthesis, we first place the study of street youth within a highly condensed history of American criminology. This history is driven in significant part by changes in the American and more recently the global economy – a force that also contributes to the problems of street youth in developed and developing countries. It is against this backdrop that social capital theory emerges as a powerful framework that synthesizes attention to socially structured background, developmental, and foreground experiences of street youth.

The depiction of social experience is most revealing when it considers people in extreme circumstances. The homeless youth who live on the streets of our cities confront desperate situations on a daily basis. Often without money, lacking shelter, hungry, and jobless, they frequently are involved in crime as onlookers, victims, and perpetrators. Yet, unlike homeless adults, who are more visible casualties of the cumulative toll of living on the street, homeless youth are difficult to distinguish from their better protected peers. Their youthfulness frequently obscures the seriousness of their problems. Fictional and media accounts rarely or sufficiently document the experiences of these youth, who are often overlooked in academic writing and research about the urban poor.

North American criminology also neglects street youth. Despite the adversity of street life and the prevalence of crime, contemporary criminology concentrates more extensively on youth living at home and attending school. This focus limits the study of more extreme social and economic situations, implying that the influence of adverse backgrounds is exaggerated and that problems of crime are sufficiently represented by the experiences of more ordinary young people.

In this book, we begin to address this neglect of street youth. We use observational, survey, and interview data gathered in two of the largest cities in Canada, Toronto and Vancouver, to document the family and school histories, living conditions, and, in particular, criminal experiences of youth who live on the streets. Our purpose is to bring into theoretical view the ways in which the background and developmental experiences of homeless youth and the foreground conditions of urban street life influence involvement in crime.

We peered further into the “black box” of street crime by introducing an indicator of “criminal opportunities” in the previous chapter. As anticipated, this variable had sizable direct effects on all types of criminal activity, as well as intervening between our foreground measure of situational adversity (i.e., nights on the street) and crime. Now we explore how criminal opportunities facilitate the embeddedness of street youth in associations that lead to the acquisition of “criminal capital.”

Broadly conceived, criminal opportunities are the situational characteristics necessary for a crime to occur. There are several recent conceptualizations of criminal opportunities, including Hindelang et al.'s (1978) lifestyle opportunity theory, Cohen and Felson's (1979) routine activities approach, and the opportunity perspective of Cornish and Clarke (1986). In general, these approaches define criminal opportunities as meetings of motivated offenders and potential victims in settings unsupervised by agents of social control.

We see a further neglected dimension to criminal opportunities. Aside from the presence of potential victims and the absence of social control agents, an important determinant of a criminal opportunity is the potential offender's definition of a situation as one suited for offending. Not all people define an unlocked vehicle, a lone pedestrian on a dimly lit street, a poorly guarded business, or an empty house as possible sites for offending. These settings only become opportunities when a person's understanding of the social world translates them into potential sites of criminality. We suggested in Chapter Four that adverse circumstances encouraged youth to interpret situations as criminal opportunities; however, other factors also stimulate this process.

Many scholars and citizens argue that we must look to individual differences as much as to social environments in order to understand the origins of social problems, including problems of the street. For example, Baum and Burnes (1993) maintain that problems such as homelessness and crime share common causes that have as much to do with the characteristics of the people involved as with the structural circumstances and social experiences they encounter. Gottfredson and Hirschi (1990) similarly suggest that a range of disvalued behaviors, from adolescents running away from home to adults engaging in force and fraud, have common origins in lapses of self-control.

The background and developmental differences that these theorists emphasize often are understood as deficiencies that derive from parentage and parenting. The psychological literature on human development encourages this orientation, suggesting that such deficiencies can lead individuals to congregate in particular environments. From this psychological or developmental perspective, problems of homelessness and crime are seen as more ontogenetic than sociogenic – that is, as being more internally than externally driven, arising more from individual differences than from variation in the social environments in which people develop (see Dannefer, 1984; Caspi et al., 1994).

Alternatively, sociologists and social psychologists who study the life course tend to focus on social structures, roles, and socialization processes. This tradition of research is inclined to view socially structured experiences as formative environments in which individual differences emerge and change, especially in the transition from adolescence to adulthood (Elder, 1975). In this and following chapters, we increasingly incorporate a life course perspective in our attention to foreground experiences of youth living on the streets.

We found in the previous chapter that embeddedness in criminal networks heightens the involvement of street youth in crime. Yet tutelage in offending is not the only dimension of street youth associations, nor are these relationships the only available source of companionship for street youth. In this chapter, we extend our analysis of street networks and explore in greater detail the associations formed by homeless youth.

Although street youth spend time alone, their life on the street is often intensely social. Research on homeless or street adults indicates that they minimize relationships with other homeless people or form short-term relationships, regardless of their intentions or the intensity of their feelings (Snow and Anderson, 1993; Fleisher, 1995). Homeless youth share some of these tendencies, but they are more inclined to enter into group relationships. Many respondents spoke about other youth with whom they “hung out,” describing how they spent their time with them and the nature of their associations. This aspect of street life is the focus of this chapter.

We explore the characteristics of street groups, the reasons for joining them, and the effects of group membership, particularly in terms of criminal involvement. Classical as well as contemporary ethnographies suggest that many youth groups are, for all intents and purposes, gangs that facilitate and promote involvement in crime. Yet, as we demonstrate later, street youth groups are diverse, offering assistance in meeting basic needs of the street, as well as facilitating involvement in crime. Moreover, several features of street youth groupings suggest that, although they share some of the characteristics commonly attributed to gangs, these street groups also have their own unique characteristics.

In chapter two, we describe the conditions that characterize homelessness and present evidence of the hardships that many youth encounter while living on the street. In Chapter Three, we demonstrate that street youth report disproportionate involvement in serious theft, an activity that we suggest partly follows from their homeless circumstances. In this chapter, we build on these findings. We explore in greater detail the effects of homelessness and argue that the destitution and desperation that characterize street life constitute important foreground class conditions that figure prominently in the genesis of street crime.

We also address a key issue of internal validity noted in the previous chapter. This concern involves the extent to which sampling street youth alone can pose problems in establishing causal processes that lead to crime. There are numerous threats to internal validity (Cook and Campbell, 1979), but the problem of restrictive sampling, or sample selection bias, often is neglected (see Berk, 1983, p. 386). In our research, the issue is whether samples of street youth constitute groups with unique attributes that might bias assessments of causal processes. Since subsequent chapters of this book are based exclusively on analyses of our Toronto and Vancouver street data, it is important to establish that explorations of such samples by themselves can validly reflect causal processes, such as the experience of destitution, that affect youth on the street. In this chapter, we present evidence that this is indeed the case.

Destitution and Crime

The idea that crime is significantly influenced by people's immediate, desperate class circumstances was once a fundamental tenet of criminology.

Many sociological theories of crime assume that adverse class conditions cause delinquency and crime. This class-crime connection is one of the central tenets of sociological criminology, from theories of anomie (e.g., Merton, 1938), social disorganization (e.g., Shaw and McKay, 1942), and subcultures (e.g., Cloward and Ohlin, 1960) to critical and Marxist traditions (e.g., Greenberg, 1981; Colvin and Pauly, 1983). Empirical evidence of this correlation is commonly found in studies of individual-level official data (see Hindelang et al., 1979; Braithwaite, 1981), as well as in studies organized around areal units, using police and court statistics aggregated into census tracts and neighborhoods (e.g., Chilton, 1964; Sampson and Groves, 1989). However, this association is often elusive in self-report surveys (Tittle et al., 1978; Weis, 1987; but see Braithwaite, 1981), leading some researchers to suggest that the effect of class on crime is weak or nonexistent and that criminologists should focus their attention elsewhere (e.g., Jensen and Thompson, 1990).

As we suggest in Chapter One, a possible source of the discrepancy between official and self-report survey findings is the latter's frequent reliance on youth who attend school. The parents of these students typically have jobs, and their children usually live at home and receive relatively good care and protection. These defining features of school criminology unnecessarily restrict variation in class origins and conditions, as well as in developmental processes through which family, school, and street experiences influence criminal involvement.

In this chapter, we adopt an alternative approach to the conceptualization, measurement, and sampling of class and crime.

This appendix summarizes our experiences studying street and school youth. We begin by describing the data collection techniques used in our original cross-sectional study. We then outline our panel study and provide details of the various strategies we used to reduce attrition. In the following section, we assess the consequences of attrition by comparing youth who were retained and lost from the panel. Finally, we discuss the validity and representativeness of our data and the suitability of using inferential statistics to analyze them.

Each of the authors of this Appendix participated in the design of at least one of the studies and was involved in one capacity or another in collecting the various data sets. Nonetheless, individuals assumed primary responsibilities for particular studies: John Hagan and Patricia Parker gathered the data for the Toronto school study and supervised the Toronto panel study; Bill McCarthy collected the data for the cross-sectional Toronto street study and supervised and took an active part in surveying and interviewing Vancouver youth for the panel study; and Jo-Ann Climenhage helped to supervise the Toronto panel survey and interviewed Toronto street respondents. We use plural pronouns in much of this discussion because they best reflect the collective nature of our work and eliminate cumbersome and awkward phrases.

Making a Start: The First Study

Our decision to study street youth arose from several conversations we had in the fall of 1986. In these discussions, we often noted the benefits of survey-oriented quantitative criminology, but we lamented the preoccupation of this approach with students, a group that lacked significant variation on key variables for the study of crime.