From sociology of deviance to sociology of risk

Journal of Criminal Justice 29 (2001) 31 ± 43

From sociology of deviance to sociology of risk Youth homelessness and the problem of empiricism Judith Bessant* School of Arts and Sciences, Australian Catholic University, Fitzroy, Victoria, Australia

Abstract This article investigates the `science of risk' and claims about its capacity to inform us about young people and the risks they present to themselves and others. To critically review the application of at-risk concepts to young people, two representative case studies are drawn on, with attention given to the ways they are informed by functionalist sociology. The discovery of the youth at risk category has largely supplanted older categories such as `delinquency' and `maladjustment' that were foundational to the sociology of deviance. Yet the methodologies, epistemological assumptions and politics of governance inherent in the older projects remain the same. Too many risk-based researches rely on normative assumptions about social and economic dependence of young people, which when given expression and legitimacy through the research findings reinforce the authority of discourses of `youth' as dependent. Many of the youth at risk researches tend to make assumptions about the category of youth as dependent and in need of close supervision. Risk-based research authorizes researchers as expert speakers about homeless youth at the same time as it delegitimates young people as speakers and active subjects capable of framing the problems in different ways. D 2001 Elsevier Science Ltd. All rights reserved.

Introduction Some sociologists may have wondered about what has happened to the `sociology of deviance' which loomed largely as one of mainstays of an older `mainstream sociology'. Some sociologists (Sumner, 1994) have even gone as far as to write an obituary for the sociology of deviance. Reports about the death of the `sociology of deviance' may, however, be premature. They should look no further than the `new' `science of risk'. The category of `youth at risk' and the associated programs now enjoy a common sense status in the minds of many social scientists, policy makers and practitioners. It has become part of the contemporary * 553 The Boulevard, East Ivanhoe, Victoria 3079, Australia. Tel.: +61-3-9499-5893; fax: +61-3-563-3605. E-mail address: [email protected] or [email protected] (J. Bessant).

common sense that leaving school `early', living in certain family arrangements and having a particular socio-economic or ethnic background put a young person `at risk' of various other social ills like unemployment, crime, suicide, homelessness, substance abuse and pregnancy. This proposition has become normative in many sectors including the health sciences, psychology, criminology and human services (i.e., youth work/ studies) (Eckersely, 1988, 1992, 1993; Potas et al., 1990; Allat & Yeandle, 1992; Haggerty et al., 1994; Hagan & McCarthy, 1997; Chamberlain & MacKenzie, 1998; Rutter et al., 1998). It has also become increasingly favoured in the education and training `industries' (Anderson, 1979; Ainley et al., 1984; Abbott-Chapman & Patterson, 1990; Dryfoos, 1990; Australian Education Council Review Committee, 1991; Bradley, 1992; Constable & Burton, 1993; Batten & Russell, 1995; Batten et al., 1991; Hawkins et al., 1995, 1998; Australian Curriculum Studies

0047-2352/01/$ ± see front matter D 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 7 - 2 3 5 2 ( 0 0 ) 0 0 0 7 4 - X

32

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43

Association, 1996; Chamberlain & MacKenzie, 1998; Sweet, 1998; Wooden, 1998). The discovery and the promotion of the `at risk' category especially in relation to young people has largely supplanted older categories such as `delinquency' and `maladjustment' that were foundational to the sociology of deviance. Yet the methodologies, epistemological assumptions and politics of governance inherent in the older projects remain the same. While the older sociology of deviance presumed the existence of a stable social order as its point of departure, the risk categories point to the prevalence of assumptions that are equally normalising about the predominance of restructuring, change and threat. Change and threat have now been tamed, as Beck (1992) suggests, by the presumption that a globalising, restructuring social formation needs to manage the multiplicity of risks it now confronts. As part of that process, `sociology of risk' has become a new way to frame old problems and preserve old projects. At the same time, there is one new and distinctive feature of the risk discourses. Like the older sociology of deviance, the primary business of the sociology of risk appears to involve, in Foucault's terms, dividing practices that distinguish between those who are at risk from certain `problems' and those who are not. While most at-risk researches give the impression of distinguishing between those at risk and `the rest', the youth at risk categories are different from older categories in terms of their capacity to incorporate the entire population of young people. Later in this article, a more detailed discussion is offered of the heritage of risk-based research (i.e., social pathology paradigms); however, before doing so, it is useful to consider the immediate socio-economic context that seems to have been, in part, responsible for having provoked much of the youth at risk research. In particular, reference is made to changes in the labor market, the impact of economic liberal policies and shifts in the economy. Young people and changes in the labor market The restructuring of the labor market and economy plus the collapse of the full-time youth labor market (Dryfoos, 1990; Langmore & Quiggen, 1994; Quiggin, 1996; Bell, 1997; Wooden, 1998, p. 35) have been accompanied by a range of specific anxieties about young people, among them youth unemployment, youth homelessness, youth suicides, juvenile delinquency and drug addiction. For the more vulnerable sections of the population such as young people, the restructured socio-economy has meant that the `youth' encounter `new morbidities'

that present major obstacles to becoming adults. Batten and Russell's position is typical: The term `at risk'. . . is used to describe or identify young people who, beset by particular difficulties and disadvantages, are thought likely to fail to achieve the development in their adolescent years that would provide a sound basis for a satisfying and fulfilling adult life. (Batten & Russell, 1995, p. 1)

This is not to suggest that the growth in the world economy and the decline in full-time secure employment have affected young people uniformly (Winefield et al., 1993; Withers & Batten, 1995). On the contrary, it is important to assess such trends on a number of levels. While recognizing that there are national and cultural differences, it is also useful to disaggregate the youth category in terms of gender, geographic location, age, socio-economic status, etc. Increased levels of unemployment, for example, tend to have a greater and a different impact on those from lower socio-economic who tend to leave school earlier than those from `the overclass' (Jencks & Mayer, 1990; Vinson, 1999). Similarly geography, ethnicity and gender matter: a decline in employment opportunities, for example, is likely to affect a sixteen-yearold Aboriginal girl living in outback Australia in very different ways than a twenty-three-year-old male English university graduate living in London. This is not to deny that there are many crosscultural and international experiences common to large numbers of young people, like increased economic insecurity, a decline in real median income, increased participation in part-time, casual and precarious employment, difficulty in financing accommodation (Rehny & McBride, 1997). While these kinds of similarities exist, there are also marked differences. In the USA, a growth in the national economy and a decline in unemployment occur in a context where social security benefits are comparatively restricted. Consequently for many young Americans, the impact of labor market restructuring has been an experience quite different from that of many young Australians living in a context where the `welfare state', although under severe attack, still offers minimal income security for jobless young people. Moreover, the US prison population of around two million, which is far more proportionately than Australia, has also meant that many young jobless Americans are likely to find themselves incarcerated. The US Gringrich-style Republican explanations of poverty that link personal immorality, family pathologies and class dysfunctions with the inability or unwillingness to work have also been popular in other `developed' countries. The international appeal and the popularity of arguments that `welfare'

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43

encourages social evils like teenage promiscuity, single mothers and social irresponsibility have been used to support a neo-conservative push to further cut what was left of the `welfare state'. Although the Australian conservative Howard government has been keen to follow the American model in respect to cutting back social security benefits, the system remains intact to the extent that most jobless young Australians are eligible for some form of government income support. The deregulation of wages in the USA has meant that for some people such as those working in lowpaid service industry, wages are so low that work becomes an economically irrational proposition. It is possible, for example, to be homeless and in dire poverty but still be working full time in the USA. Despite the fact that conservative Howard government in Australia makes no secret of their determination to lower wages of some workers and to see the unemployed hired at a discount wage, the idea of `the living wage' continues to protect low-paid employees by preventing their earnings from dropping below that which can sustain a minimum standard of living. But Australia does have `a youth wage' which quite significantly reduces the income for most people under twenty-one years of age. Despite this, low wages have not been identified as a disincentive to work as it has been in the US (Devine & Wright, 1993). The decline in the USA of smokestack industries, the increased employment in the new service economy and the emergence of information industries have had a significant impact on young workers and `the working poor' (Sassan, 1991; Bauman, 1998). Another point of difference between the US, the UK and Australia is evident in a talk about a juvenile underclass inherent in USA and UK debates about poverty and `the poor' (Devine & Wright, 1993; Gans, 1995; Wilson, 1996; McDonald, 1997; Shragge, 1997). In recent years, Australians have had warnings offered by various social critics about the emergence of an American-style juvenile underclass (Watson, 1993, pp. 14 ± 16; White, 1994; Bessant, 1995, pp. 32 ± 48). Such tends to be promoted with little regard to the enormous cultural, legal and social differences between the USA and Australia. Those differences include everything from large variances in the ethnic composition of both countries to the radically divergent legislation and cultures relating to the use of firearms. There are also major differences in the demography and density of our respective populations. Australia simply does not have large cities like Chicago, New York, Detroit, Philadelphia, New Orleans, Atlanta, Houston, Birmingham, San Antonio, Miami and Oakland. Australian cities are

33

small on an American scale which also makes a difference in terms of the emergence of a `juvenile underclass'. As was just indicated, most western contemporary societies are undergoing quite fundamental changes that contest enlightenment-based modernity and open dominions where people decide on new forms of the political and social. One way that those who are participating in social scientific debates have sought to understand these new arrangements is through the idea of risk (Beck, 1999).

Knowledge about youth at risk There are two ways that risk can be known. The first is by direct experience such as the knowledge of danger people develop through immediate encounters (i.e., the risk of unemployment and/or accidents). The second way risks can be known depends on the research and knowledge of `experts' who constitute a `culture of scientism'. As Beck (1992, p. 3) argues, this culture of scientism is central to `risk society' because: . . . the consequences of scientific and industrial development [are] a set of risks and hazards which are no longer limited in time and space and for which no one can be held accountable. Although many pronouncements made by experts are also voiced without a direct experiential or empirical referent, they nevertheless draw on the authority of the expert as the knowledge maker for their credibility.

Knowledge about `youth at risk' is usually not gained through our immediate experience, but rather through investigations that constitute the knowledge of experts. For writers like Foucault, the means used to discover, verify and manage risk to social order have always been a deeply political and moral process (Foucault, 1977; Holloway & Jefferson, 1997, p. 258). It is a process designed to change uncertainties into probabilities, thereby making them amenable to impersonal administrative regulation based on `scientific principles'. This article offers an investigation of current projects directed towards identifying young people alleged to be `at risk' of becoming victims and/or threats to social order. There are a number of projects committed to identifying `youth at risk' including youth homelessness, youth unemployment, substance abuse, juvenile crime and youth suicide. The application of scientific methodology makes possible both a scholarly discipline, and in Catalano's (1998) case, a `science of prevention'. The research used to identify the various `at risk' populations is closely linked to the policing of these problems especially through `the commu-

34

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43

nity', the schooling system and `the family' and includes a range of preventative programs designed to stop youth unemployment, homelessness, juvenile crime and youth suicide. As mentioned, risk-based research is part of a long heritage; its immediate predecessor Ð the sociology of deviance Ð itself stood on the shoulders of longstanding and highly influential tradition of criminological and urban anthropology. Risk-based research: the legacy of a long tradition To appreciate this tradition, we need only look at the Chicago School of Sociology (re: Park, 1915; Thomas, 1923; Thrasher, 1927; Shaw & McKay, 1931) to see how much of their empirical and theoretical focus touched directly on the issue of anti-social behaviour and the use of urban space. The American universities of the time also stressed the practice of applied and theoretical research and the training of professionals, and as they adopted the cult of entrepreneuralism, academics promoted the idea of a practical intelligence (Bledstein, 1976). Moreover, the political and social organization of knowledge production and certification of legitimate knowledge characterised a `culture of professionalism'. Between 1895, when Nordau (1895) published his Degeneration, and 1915, when John Gillin published Social Pathology, there was a floodtide of descriptive social problems and, in particular, criminological research. Still identified as part of the American progressivism, the growing emphasis on social problems reflected many American preoccupations with causality. Increasing numbers of projects were devoted to identifying the connections between crime and factors such as homelessness, poverty, alcoholism, family life Ð particularly in relation to immigrants, `Negroes', First Nation people, lunatics, imbeciles and the `racially unfit'. Although these progressives identified biological factors as important, they did not discount the role of social and cultural factors, thereby opening a rich field for professional disputation and debate. Hooton's (1939) The American Criminalspecified the biological inferiority of criminals, while (later) Glueck and Glueck's (1965) Physique and Delinquency made the same point (Bessant et al., forthcoming). This `respectable anxiety' of early 1900s with crime and poverty drew on the sensibilities of progressive urban anthropologists as well as child savers worried about the plight of children of `the poor'. The focus had much to do with the social problems of pauperism, those who lacked housing and the influence of urban space. As early as 1923, W.I. Thomas

researched the lives of young female prostitutes working in Chicago resulting in the publication of The Unadjusted Girl. Anderson (1923) from the Chicago School produced The Hobo in which he described the predicament of homeless men and petty crime. In 1927, Thrasher produced the book The Gang and Whyte's (1943) work resulted in the publication Street Corner Society. Two years later, Sutherland and Lock (1936) published Twenty Thousand Homeless Men. Spanning the centuries, we have witnessed many attempts to categorize and count the number of the `vagabonds' and criminals. Dugdale (1877) published a study of the Jukes, an `all-American family' of criminals and paupers. Committed to a method of scientific research which linked historical ± biographical syntheses to statistical analysis, Dugdale's intention was to demonstrate the cumulative effects of heredity and social environment in the manufacture of criminality. An initial study found that of the twenty-nine Juke family males, aged from fifteen to seventy-five, seventeen were `criminals'. A more extended study of 709 Jukes, dead and alive, found that 76 had committed crimes and 180 had received `poor relief' (or social security). A hundred years on, Marvin et al. (1972), reporting on their longitudinal study of 9945 boys born in Philadelphia in 1945, found that some 3,475 of the boys had been involved with police at least once. While not originally designed to be etiological (i.e., designed to establish the causes of the problem), Marvin et al. found that while 71 percent was comprised of whites and 29 percent non-whites, non-whites were more likely to offend and were far more likely to re-offend. Their research suggested that a combination of being non-white, failure to graduate from high school, `weak IQ', repeated intra-city migration and membership of `lower socio-economic groups' was the major risk factor pointing to those most likely to engage in anti-social conduct (1972, pp. 244 ± 255). Risk-based research is part of a long history of popular and scholarly concern about social and historical decline. Historians like Pearson (1983) demonstrated how each generation believes itself uniquely threatened by the signs and symptoms of decay and degeneration often signified by concerns about lawbreakers, vagrancy and the threat of hooligans. This raises the question of whether the present talk of risk and young people ought to be seen as just the latest expression of a long tradition of fear and anxiety. Such an argument, however, is not particularly useful. It could be argued that the contemporary preoccupation with risk indicates that our lives are now either less secure or less oriented to hopeful possibilities than at any other point in history. It may

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43

be very difficult, if not impossible, to establish a rational basis for evaluating this. Historians like Herman (1997) point out that cultural pessimism is a condition that is not unique to western societies at the end of the 20th century. Pick (1993) also observes that adults have disapproved of young people's perceived depravities since Socrates complained about Athenian youth. It is more than a coincidence that the modern talk of youth at risk has emerged at a time when the public policy and the public culture of so many western countries have been reshaped by a resurgence of liberal individualism. Without reducing the emergence of talk about risk to an expression of resurgent liberalism, it seems that there are some affinities between the two. Resurgent liberalism takes its form variously in a near-hegemonic policy discourse grounded in neo-classical economics and preoccupied with individual choices and freedoms and intent on re-working activities in the public sector via marketbased activities and metaphors. The actuarial tendencies of economic liberalism, which equate rationality with calculation, are paralleled in social scientific research which calibrates risk factors and claims to predict the degree to which a particular kind of person is `at risk'. In this article, what are seen as the essential elements of a `science of risk' are identified and consideration is given their capacity to inform us about young people and the risks they are said to present to themselves and others. This is achieved through a study of youth homelessness and a discussion of some of the problems associated with the empiricism central to the identification of youth at risk of either being or becoming homeless. It is argued that there is good reason not to accept the proposition that the `empirical' research said to characterize risk identification is credible simply because it reports in an objective fashion what is actually there. Homeless young people and the problem of empiricism In Australia, like most other Western countries, we have seen the development and implementation of projects directed towards identifying those at risk (and even those `potentially at risk') of homelessness (Neil & Fopp, 1992; Cordray & Pion, 1997; Hagan & McCarthy, 1997; Chamberlain & MacKenzie, 1998). The primary purpose of such projects is to provide a basis for effective intervention. In Australia, one research project (Chamberlain & MacKenzie, 1998) offers a useful basis with which to assess the value of the `empirical research prac-

35

tice' that underpins most risk researches. This will be analysed in conjunction with a Canadian study of homeless young people in Toronto and Vancouver (Hagan & McCarthy, 1997). Based on a census that elicited responses from 99 percent of secondary schools, Chamberlain and MacKenzie (1998, p. ix) implemented a survey of 42,000 young people to identify those `at risk' and `to identify policies and practices that enable early intervention' to prevent youth homelessness. The research of Hagan and McCarthy, on the other hand, was built around two studies: a 1987 ± 1988 cross-sectional comparative analysis of young people on the streets and in schools in Toronto, and a 1992 panel study of `street youth' in Toronto and Vancouver (Hagan & McCarthy, 1997, p. 1221). From the start in each project, we are invited into the authoritative and verifiable world of the modern social sciences. The first sentence in Chamberlain and MacKenzie's text informs us that: `This book is the result of a research journey which has taken eight years' (1998, p. vii). The assured voice of modern social science is present in their conclusion that: On the basis of these findings, it is possible to make generalizations about the `at risk' population in most communities. In a typical city school with 1000 students, there will probably be about 100 to 140 young people (10 ± 14%) who are possibly at risk at any point in time, and this will include 40 to 60 students (46%) who are seriously at risk. The latter group are likely to be experiencing major problems in family relations. Most will not be happy at home, many will feel unsafe, and some will be running away. (1998, p. 98)

Evident also is a scattering of tables throughout the Chamberlain and MacKenzie book, conveying a considerable amount of numerical information. Thus, the reader is assured that the product of this research will provide a credible basis for identifying `youth at risk'. The reader's confidence is justified because both projects rely on legitimate social scientific methods, i.e., empirical surveys of young people `at risk' of homelessness. The primary interest of Hagan and McCarthy is criminological; as they argue ``. . . street youth are not a significant focus for contemporary criminology. This book is an argument for returning to the street and renewing attention to its implications for understanding crime'' (Hagan & McCarthy, 1997, p. 21). Hagan and McCarthy are interested in `peering' into the `black box' of street crime to present `indicators of criminal opportunities'. Such variables, they argue, have a sizable and direct effect on all types of criminal activity and will therefore enable them to ``intervene between our foreground measure of situational

36

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43

adversity (i.e., nights on the streets) and crime'' (Hagan & McCarthy, 1997, p. 135). Although Hagan and McCarthy rely more on interviews than their Australian counterparts, like the latter researchers, they operate with specific views about what knowledge and truth is and how they can be scientifically secured via naturalistic methods. Hagan and McCarthy declare this, themselves, when they explain how they depend for theoretical framework on theories like and Cloward and Ohlin's strain, control and differential association (Hagan & McCarthy, 1997). Strain theorists claim that poverty and inequality produce crime. Cloward and Ohlin (1960) argued that young people who could not obtain material wealth by legitimate means turned instead to illegitimate methods involving criminal activities. They asserted that one method of coping with the stresses and frustrations of being thwarted in attempts to achieve `success' is to develop alternative avenues of success like crime. Strain theory also shares some commonality with `differential association theory' that relies on the assumption that delinquency is passed on by associations with `bad' company. This means that if there are many street-based, poor, unemployed young people, they are likely to `infect' others and create a culture of crime. Insistence on drawing these causal connections encourages a neglect of other possible explanations for conduct, such as leaving home `early', or crime Ð besides poverty, resistance or detachment. Not only does this deny the agency of young people, it ignores a vast literature that has been produced since the 1960s which questions the claim that the action of young people can be understood in terms of constraints and structural imperatives and claims that homelessness can be adequately explained through strain and `differential association theory'. Equally problematic are propositions that it can be `obvious' or `empirical', that a particular person is `at risk' of becoming homeless, criminal or engaged on illicit drugtaking. Such a proposition depends on a specific kind of research, i.e., epidemiological research, which attempts to identify at risk factors for a large range of social or health problems.1 Rigorous epidemiological research involves establishing that: . . . an aspect of personal behavior or lifestyle, an environmental exposure, or an inborn or inherited characteristic. . . on the basis of epidemiological evidence is known to be associated with healthrelated [or social] condition[s] considered important to prevent. (Last, 1988, pp. 115 ± 116)

Unfortunately, the research performed by Chamberlain and MacKenzie or Hagan and McCarthy do not attempt to meet, let alone pass, this test. More-

over, the proposition that one can use aggregate data about large numbers of people and then apply that data or any finding from it to a particular person and from there go on to argue that the individual is `at risk' involves being quite unempirical. As Gould argues, it relies on a peculiar jump in logic. As a statistician, Gould (1996, p. 3) reminds us that: . . . reality is composed of varying individuals in populations and that variation itself is irreducible.

The reasoning used in this exercise involves taking from the epidemiological research which may show certain average values or deviations from the norm (based on investigations of large numbers of individual cases) and then turning to an actual single individual and saying to that person `Because you exhibit factors a, b and c, you are at risk of X'. Such an assessment means moving from measures of central tendency, like averages, to particular cases. As Gould explains, ``Central tendency is an abstraction, variation on the reality'' (1996, pp. 48 ± 49). In other words, moving from a claim that X is true of the whole group to the claim that X is also true for each single member of the group cannot be done. Yet even these useful preliminary warnings do not fully prepare us for what Chamberlain and MacKenzie (1998) offer. Conceptual problems The conceptual confusion in which Chamberlain and MacKenzie operate with is revealed early. They acknowledge (1998, pp. 16 ± 21) that: 

there are many different definitions of `homelessness';  that some writers have argued there are no correct definitions and that the concept is arbitrary, not very helpful and/or should be abandoned;  that certain definitions of homelessness, which depend on the `perceptions' of young people, constitute an `extreme form of relativism';  that this `relativism' can . . . be overcome theoretically once it is recognized that `homelessness' and `inadequate housing' are socially constructed, cultural concepts that only make sense in a particular community at a given historical period. . . it is a cultural construct, but this does not mean that `homelessness' is just a matter of opinion, or that all definitions are `arbitrary'. (1998, p. 19) (authors' stress)

Hagan and McCarthy, on the other hand, do not bother with such exercises in clarification of homelessness except to specify that the term youth is ``roughly between the ages of fifteen and twenty

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43

four'' (op. cit., p. 11). Homelessness or precisely what counts as `living on the street' is not made clear. The closest they come to specifying homelessness is through an inventory of other people's research where they describe how those writers tended to ``conceptualize narrowly such youth as runaways''. . . (ibid, p. 7). In the case of Hagan and McCarthy, we can only assume that the category was so obvious to them that it did not require explanation. More generally, when social scientists acknowledge that there is considerable controversy about the meaning of a concept, we should not expect to encounter a simple empirical exercise of describing or measuring something that exists (like the number of marbles in a bag). Ideas and social experiences like poverty, homelessness and unemployment are complex, multi-faceted and do not exist in the same ways that, for example, a marble or a bag exists. Instead, they depend on social definitions that there may be little or no consensus on. Accepting this seems to jar the confidence of some social scientists, creating an uncertainty about aspects of their research program, and Chamberlain and MacKenzie (1998) appear to be no exception to this unease. Chamberlain and MacKenzie are eager for the security of objectivity. Their achievement of this is analogous to discussions on the problems involved in researching other social science categories like `poverty'. Confronted with considerable differences about what `homelessness' or what `adequate housing' means, they resolve the puzzle in a `classic' way. MacKenzie and Chamberlain maintain that an objective `community standard' exists which reflects the standards of a particular culture or society regarding to `homelessness'. Thus, they claim that: . . . community standards are usually embedded in the housing practices of a society. These identify the conventions and cultural expectations of a community in an objective sense. . . (1998, p. 19)

This enables the social scientist to empirically gather the data to demonstrate `objectively' what homelessness is, and this is achieved by identifying anything which falls below `the standard'. This is problematic because it depends on claims about the existence of a common, but not particularly useful, idea like: (i) the notion [that] a consensus exists in a `society' based on `shared community standards' `according to [which] the conventions and expectations of a particular culture' can be specified about what in this case constitutes `adequate housing' or `homelessness' (Chamberlain & MacKenzie, 1998);

37

(ii) This claim depends on a prejudice shared by many sociologists about the concept of a consensual and singular `social system' called `society' or `culture'. This notion flies in the face of the reality that societies like ours are marked by extreme inequalities in the distribution of income and wealth, consumption of goods and services. This is not to mention immense cultural differences and variances in expectations and lifestyle. Thus, any claim that there is either a convention or a consensus about adequacy disintegrates or is inherently problematic. Chamberlain and MacKenzie's do not seek to construct an index of homelessness; they simply (1998, p. 20) assert that there is `benchmark' or a `minimum community standard', an assertion that is not demonstrated. For social scientists like this, `research methodology' offers some protection from any danger to which the dreaded `conceptual relativism' might expose them. Central to the faith of the `empirical' social sciences is the proposition that unless the concepts, entities or objects that a discipline seeks to know are actually there (phenomenalism), are amenable to measurement and display an innate orderliness and predictability, then they do not exist and/or have no place in the schema of assured knowledge. Principles of objectivity and the magic kiss The critical retreat for social scientists who adhere to `the principles' of objectivity is the proposition that scientific method, in some way, authorizes all the claims made by scientists. `Scientific method' legitimizes certain styles of understanding above others. It depends on measurement, replicability, experimental procedures designed to isolate causal relationships and the generation of explanatory, predictive statements which are recognized as the hallmarks of `scientific method'. In terms of `scientific practice' principles such as naturalism, objectivism and operationalism operate as authorized directives for good research practice. Danziger identified this framework as the Sleeping Beauty model (1990, p. 2). According to this model, the objects that science operates with or seeks to `know' already exist, and only await the magic awakening kiss of the scientific researcher to animate them and bring them into the province of knowledge. Practitioners of this model believe that objects already exist, and because they are orderly and actual entities means they can be measured, tested and described accurately, then subjected to experimental (repeatable) procedures, for the purpose of establishing invariant

38

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43

relationships of causality and/or probability. However, for entities that do not have these qualities (like `at riskness'), the processes of operationalizing them means that as abstract entities or non-tangible phenomenon, they can nonetheless be rendered amenable to measurement, testing, description or experiment for the purpose of establishing relations of co-variance/invariance, etc. Most social scientific works depend on practices such as operationalization, representation, measurement and repetition of those practices. These are practices that are discursive and social in nature: The sentences in textbooks, the tables and figures in research reports, the patterned activities in research laboratories are first of all products of human construction whatever else they may be as well. (Danziger, 1990, p. 2)

Though this insight is resisted by empiricists (and positivists), the `natural attitude' maintains that `reality' or `nature' is simply there awaiting `the magic kiss' of scientism (Danziger, 1990). All the sciences work with constructive schemes which mandate and regulate certain assumptions about what is real and what counts as knowledge as well as a range of `scientific practices' for which there may be no empirical or rational justification. Operationalizing the category or model Operationalism involves using a category for which there is no actual empirical referent (like `homelessness' or `at riskness'). It involves establishing criteria that allow the research process to develop and defines the phenomenon to be `discovered' and then `measured'. As Chamberlain and MacKenzie (1998) explain, the category of `at risk' helps to grasp the otherwise elusive concept of `at risk'. The reader also learns that any `at risk' index needs to be identified numerically ``because it is necessary to make a quantitative assessment of the at risk population. . .'' (1998, p. 89). Evident here is the enactment of a standard inconsistency by empiricist researchers who are concerned with `objectivity' and the need to demonstrate that their work is `empirical'. This is despite the reality that the category of `at risk' is not empirical. Yet, this seems to be a relatively minor problem given that records, tables, indexes and a variety of measurement processes are there to convince the laity otherwise. How did they do this? They explain: Our task was to design a survey to be filled out by secondary students in a classroom situation which could identify young people who might be at risk. In

order to do this, it was necessary to operationalize the concept of `at risk' in a simple way, and to develop questions that would be easily comprehensible to students from years 7 to 12. (1998, p. 91)

Perhaps it can be said in Chamberlain and MacKenzie's favor that unlike their North American counterparts, they did at least demonstrate an awareness of the need, according to their own intellectual framework, to articulate the categories they were examining. All these raise the question about how such researchers establish the `risk factors' and thus the substance of the questions? Chamberlain and MacKenzie do not say how explicitly, but it seems they used the `judgments' of certain welfare teachers at schools about what they saw as risk factors. They refer to this when explaining that `our operational definition of `at risk' is grounded in `the first order experience of these workers' (1998, p. 90). These `judgments' are said to be far better than the `perceptions' of the young people they criticize as so `relativistic'. Given that Hagan and McCarthy are interested in young people already on the street to determine the likelihood of them engaging in criminal activities, we see a different set of indicators emerge (i.e., employment status of parent, whether the family is `intact', `erratic parenting', `explosive parenting', school involvement and involvement in theft) (ibid, pp. 63 ± 64). Most of these `risk factors' allegedly related to `the family situation'. Using the `risk factors' identified and with access to `a complex body of qualitative information', and the fact that many months, if not years, of data collection and analysis have been invested in Chamberlain and MacKenzie's project, the social scientists are well situated to construct a survey instrument using five questions scored from zero (no risk) to two (at risk) which asked things like: `Have you run away from home in the past 12 months?' or `Do you feel safe at home?' (Chamberlain & MacKenzie, 1998). Even if the most simple criteria are used, namely that there has been an attempt to articulate what the links are between the alleged `at risk' factors ``known to be associated with health-related [or social] condition[s] considered important to prevent'', Chamberlain and MacKenzie (1998) have not been successful. They do not establish any link that could support their claims that they identified a population of `at risk' young people. For Chamberlain and MacKenzie, the risk factors are to be made known by asking questions like ``. . . whether the young person has. . . run away from home in the past twelve months?'' (1998, p. 91). (Running away, they argue, is `usually a sign

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43

of serious problems at home'.) The researchers simply assert that: In most cases where a young person is at risk, there is a serious problem in family relations. (1998, p. 91)

This statement is made without any justification or evidence, nor is there an indication given about whether or not those young people who reported these circumstances did actually leave home. What any of this actually refers to, or why it should be regarded as convincing, is not made clear. They have, in one act of omission, ignored the central requirement (Last, 1988, pp. 115 ± 116) involved Ð identifying a `risk factor' involves establishing some aspect of personal behavior or lifestyle, an environmental exposure, or an inborn or inherited characteristic Ð which, on the basis of epidemiological evidence, can be associated with health-related [or social] conditions. The next step in securing objectivity (i.e., measurement) is to assign a particular risk factor a numerical weight so that the overall degree of `at riskness' can be `accurately determined'. As Chamberlain and MacKenzie explain: A young person who had run away scored two, and those who had not scored zero. The second question asked. . . whether they felt `safe at home'. Everyone who did not feel safe scored two. (Chamberlain & MacKenzie, 1998, p. 91)

Other questions to be asked include whether they `get into a lot of conflict' with parents and would they like to move out soon, or whether they `feel happy at home' or not. Scores are kept from naught, one or two depending on the answer (Chamberlain & MacKenzie, 1998, p. 91). `Taken together (the reader is told), these five questions provide an indicator of current family circumstances' and thus the level of risk of homelessness the young person faces (Chamberlain & MacKenzie, 1998, p. 91). In their analysis, MacKenzie and Chamberlain subdivided the youth cohort into country regions, major regional city areas, middle class, traditional working class, and a `new working class' (located in the corridor outskirts of the city), Non-English speaking background families, Anglo-Australians and Australian aboriginals (1998, pp. 92 ± 97). This means that the proportion of those most at risk can be developed and the risk factors of children from specific backgrounds more readily identified by reference to their ethnic or socio-economic status. If, for example, you are seen as belonging to the `new' or `traditional' working class, then you are measured as being at a far higher risk of being homeless than a young person from a `middle class' background (1998, p. 95).

39

Not only are there problems with many of the assumptions underlying the questions asked, but there are also difficulties with the assumptions underlying the categories used. For categories like `middle class' or `traditional working class' to be operationalized assumes the existence of actually homogenous groups whose experiences can be isolated from other groups/classes and whose members actually do share characteristics. Inherent in such categories is an assumption about the existence of a homogeneity, general essence or trait shared by all members of the specified group. While it is true that many groups of people share some specific experiences, there is a question whether we can extend those quite specific commonalities to talk about a `middle class' or `traditional working class' or `rural' young people. In this article, the point has been made that claims by some representative empirical social scientists that they have established an `objective' or an `empirical' basis for identifying a population of young people who are `at risk' need to be seriously questioned. Chamberlain and MacKenzie (1998) and Hagan and McCarthy offer a good examples of why these claims should be disputed. Risk-based policy and governance Scientific endeavours to regulate `youth at risk' are not new. Historically, the `naturalist' or positivist' agenda have long informed research and been used to identify forms of social management strategies. Thus, in a context of limited public resources and the push for greater efficiency, it is not surprising to see project developed to `target' those most readily identified as being `at risk' or `potentially at risk'. By targeting youth at risk, `the state' is able to implement strategies that directly relate to the greatest danger of harming themselves and/or others. This is achieved through a range of interventions informed by disciplinary notions of treatment, rehabilitation, reform and integration (Hawkins & Weis, 1985; Catalano & Hawkins, 1996; Resnick et al., 1996, pp. 5 ± 8). As the youth at risk projects provide for more specific `targeting', so can interventions be more sharply directed. Questioning the assumptions underlying these projects Ð that the individual young person (or group of people) is central to `the problems' (of youth unemployment, pregnancy, homelessness, leaving school `early', etc.) Ð plus a refusal to focus on `the causes' of `the dangers' or `risk' divert attention away from the disciplinary technologies of power as they now operate. They also divert attention from the role of those involved in the research and service enterprises whose livelihood has come to depend on

40

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43

the constitution of those problems. This is not to ignore the plight of young people in various forms of distress; it is simply to argue that in understanding `youth at risk', it is critical not to overlook the powerful role of the adults involved in researching and responding to the problem. While there are obvious differences between the various youth problems (i.e., homelessness, suicide, crime, unemployment, etc.) and slight variations in methodologies used to detect the risk factors, they are all directed towards the dispersal and extension of governance over a specific section of the population, namely young people. Moreover, all emphasize regulatory management and coercive intervention. One powerful selling point deployed by advocates of the youth at risk projects has been the claim that they offer a more efficient form of disciplinary power. For risk experts like Catalano, this efficiency operates both economically and politically. It involves `preventing the problems before they occur' which is said to be a cheap way of fixing otherwise expensive problems (Catalano, 1998). Prevention is presented as a cost-saving measure which is quite appealing for governments strapped for cash and keen for an image of responsible fiscal management. Youth at risk experts frequently talk in terms of cost benefits analysis as they employ persuasive techniques like measurement and weight metaphors in publicizing the virtues of their programs Ð in Catalano's (1998) case, ``a 100 grams of prevention is worth a kilo of treatment''. Moreover, they are very effective because they key into progressive ideals about the virtues of local autonomy and `community'. This, in turn, frequently dovetails with conservative agenda which are translated into moral arguments about a `need' for greater individual responsibility as a means of solving those problems (re: the `blame the victim' approach). Politically, new risk management and preventative programs offer greater efficiency in terms of governance because of their relatively low cost compared with resource-hungry `treatment' models. Politically, they promise discretion, `relatively low visibility' and thus a reduced likelihood of resistance. They are efficient also because of their potentially enormous reach. In other words, the preventative, corrective or ameliorative models of youth at risk research are highly attractive politically because they have the capacity to pervade all aspects of social life and replace more punitive regimes of discipline with risk-based technologies of governance. This includes a number of connected changes: the demise of socialized social insurance (i.e., welfare) which includes managing groups like the unemployed, and

the poor. . . to its replacement with the privatized components of what was the public sector in conjunction with risk-based techniques of governance that allows for a more market-based and prudent use of state resources to manage the population (O'Malley, 1992; Rose, 1996). The political programs that develop from the atrisk research enterprises focus upon doing something about `practical objects' like the reduction of levels of unemployment, rates of crime, youth suicide or youth homelessness, These are projects that provide recipes ``for corrective intervention and redirection. . .'' (O'Malley, 1992, p. 258). Rose (1994) argues that market mechanisms are used increasingly to link the `active individual' and family to the expert. Moreover, active policies, he argues, are set in place which allow for few opportunities to contest expert authority. Through youth at risk research, project policies can be directed specifically to those `at greatest risk' and consequently `in greatest need of intervention'. Young people calculated as posing the greatest risk to themselves and their community due to their failure to manage themselves are the `high-risk groups' `targeted' for government and professional responses. The ascendancy of conservative economic liberalism can be observed as the moral high ground is taken in claims that welfare `saps the enterprise of the individual' and restricts the operation of market forces. Being active and competent in terms of selfmanagement is part of the rhetoric put forward by contemporary conservative interests in the push towards the implementation of more economic liberalists policies. Self-management, being responsible for yourself and subject to the market forces, is said to restore individual moral standing. In promoting their professional interests, expert intervention directed toward increasing the responsibility of individuals by improving their capacity for self-management is described as a form of empowerment. What can be observed are new forms of expert practices which are by no means universal, but are spreading rapidly. The role of the expert is clear: to intervene when the `client' is recognized as being deficient in terms of their ethical, emotional, social, intellectual abilities necessary for self-management. This activity is not confined to one discipline or professional practice; rather, what can be seen is an extension of this approach into a multitude of disciplines and practices that facilitated the logic and practices of a system of `seamless services' and `onestop' institutions that service all the `needs' of young people at risk (Dryfoos, 1990, 1994, 1996; Department of Human Service Western Metropolitan Region, 1998).

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43

Conclusion Most contemporary human service professional practices and policies are shaped by a framing of the problem reliant on risk discourses. The evolution of policies and programs centered on the category of `youth at risk' has extended the reach of governance in youth-based institutions like the schooling system. Dryfoos, for example, offers a model of what she calls a `one-stop' service working through the education system offering a variety of human services directed to addressing aspects of one's life. Given that all young people are legally obligated to attend school, the reach of risk-based enterprises and thereby the extension of the scope of governance of young people are enormous. The youth at risk category has largely replaced older classifications of deviance like `delinquency' and `maladjustment' that were foundational to older models of education and welfare practice that drew their inspiration and categories from a certain kind of sociology. Despite decades of critique of structural functionalist social science relying on positivist techniques, the tradition continues to live on through new risk research. Risk-based research has a rich heritage to draw on in `mainstream sociology', as well as on popular anxiety about `social problems' like vagrancy, the poor, crime and hooliganism. Equally the immediate socio-economic context marked by the restructuring of the economy and state under the aegis of economic liberalism has, in part, been responsible for sustaining much of the youth at risk research. The novel feature, however, is that the risk category has the capacity to embrace the entire youth population, far exceeding the scope of older practices informed by deviancy theory. To critically review the application of at risk concepts to young people, two representative case studies were drawn on, with attention given to the ways they are informed by functionalist sociology, and the problems associated with efforts to achieve objectivity and to identify `causes'. It was indicated how risk-based research connects with the governance of young people, and how risk researchers tend to locate the source of the problem within the individual, their family and immediate community. It was argued that the projects deeply flawed ontologically and epistemologically, producing outcomes that are not practicable nor desirable. Moreover, the current dominance of such approaches means that alternative framings and arguments about the sources of various youth problems receive comparatively little attention. Too much risk-based research relies on normative assumptions about the social and economic dependence of young people which when given expression

41

and legitimacy through the research findings reinforce discourses about `youth' as dependent. Much of this work tends to rest on assumptions about `youth' as dependent and in need of close supervision. Riskbased research authorizes researchers as expert speakers about homeless youth at the same time as it delegitimates young people as speakers and active subjects capable of framing the problem in different ways. If problems like youth at risk of homelessness were constituted differently, e.g., as the product of a failure on the part of communities to recognize the capacity and entitlement of young people to access independent living and secure employment, then it is likely that we would end up with very different policies from those that currently exist. Notes 1. Different types of research operate within the epidemiological research program. They include: 1. `Case studies' which research on people who have a particular problem; 2. `Correlation studies' which survey the entire population for correlations between factors a, b and c and the problem; 3. `Cross-sectional surveys' which sample surveys a complete population; or 4. a `Cohort study' which selects a part of a population by age, ethnicity, etc.

References Abbott-Chapman, J., & Patterson, C. (1990). Evaluation of the students at risk program. In: Tasmanian Schools, Report 1 (pp. 27 ± 34). Hobart: Youth Education Studies Center. Ainley, J., Batten, L., & Miller, H. (1984). Patterns of Retention in Australian Government Schools. Hawthorn: Australian Council for Educational Research (Research Monograph No. 27). Allat, C., & Yeandle, S. (1992). Youth Unemployment and the Family: Voices of Disordered Times. London: Routledge. Anderson, D. (1979). A social ± psychological theory of school drop outs: an exploratory study. In: G. Rowley (Ed.), Proceedings of the 1979 Annual Conference of AARE ( pp. 282 ± 294). Melbourne: AARE. Anderson, N. (1923). The Hobo: The Sociology of Homeless Men. Chicago: University of Chicago Press. Australian Curriculum Studies Association. (1996). From Alienation to Engagement (Vol. 1 ± 3, pp. 10 ± 24). Canberra: ACSA. Australian Education Council Review Committee. (1991). Young People's Participation in Post-Compulsory Schooling: Report of the Australian Reduction Council Review Committee. Canberra: Australian Government Publishing Service (Finn Report).

42

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43

Batten, M., & Russell, J. (1995). Students at Risk: A Review of Australian Literature 1980 ± 1994. Hawthorn: Australian Council for Educational Research. Batten, M., Withers, G., Thomas, C., & McCurry, D. (1991). Senior Students Now: The Challenges of Retention. Hawthorn: Australian Council for Educational Research. Bauman, Z. (1998). Work, Consumption and the New Poor. Buckingham: Open University Press. Beck, U. (1992). Risk Society: Towards a New Modernity. London: Sage. Beck, U. (1999). World Risk. Cambridge: Polity Press. Bell, S. (1997). Ungoverning the Economy. Melbourne: Oxford Univ. Press. Bessant, J. (1995). The discovery of an Australian `juvenile underclass'. Aust N Z J Sociol 31, 32 ± 48. Bessant, J., Hil, R., & Watts, R. (forthcoming). Understanding Criminology. Cambridge: Polity Press. Bledstein, B. (1976). The Culture of Professionalism. New York: W.W. Norton. Bradley, G. (1992). Increasing student retention rates. Youth Stud Aust 11, 37 ± 42. Catalano, R. (1998). Young people in Communities Forum, Hosted by the Victorian Department of Human Service, Melbourne. Catalano, R. F., & Hawkins, J. D. (1996). The social development model: a theory of anti-social behaviour. In: J. D. Hawkins (Ed.), Delinquency and Crime: Current Theories ( pp. 149 ± 197). New York: Cambridge Univ. Press. Chamberlain, C., & MacKenzie, D. (1998). Youth Homelessness: Early Intervention and Prevention. Sydney: Australian Centre for Equity Through Education. Cloward, R., & Ohlin, L. (1960). Delinquency and Opportunity: A Theory of Delinquent Gangs. New York: The Free Press. Constable, E., & Burton, K. (1993). Prevention and politics: giving at-risk youth a future. In: D. Evans, M. Myhill & J. Izard (Eds.), Students Behaviour Problem: Positive Initiative and New Frontiers ( pp. 16 ± 24). Hawthorn: ACER. Cordray, D. S., & Pion, G. M. (1997). What's behind the numbers? Definitional issues in counting the homeless. In: D. Culhane & S. Hornburg (Eds.), Understanding Homelessness: New Policy and Research Perspectives (pp. 14 ± 23). Washington, DC: Fannie Mae Foundation. Danziger, K. (1990). Constructing the Subject: Historical Origins of Psychological Research. Cambridge, UK: Cambridge Univ. Press. Department of Human Service Western Metropolitan Region (1998). Service context and intervention approach of the school-focused youth services. Victoria, Preliminary Discussion Paper No. 1. Devine, J., & Wright, J. (1993). The Greatest Evil: Urban Poverty and the American Underclass. New York: Aldine Gruyter. Dryfoos, J. (1990). Adolescence at Risk: Prevalence and Prevention. New York: Oxford Univ. Press. Dryfoos, J. (1994). Full Service Schools: A Revolution in Health and Social Services for Children. San Francisco: Youth and Families.

Dryfoos, J. (1996). Full service schools. In: Educational Leadership ( pp. 16 ± 20) (April). Melbourne:VUT. Dugdale, R. (1877). The Jukes: A Study in Crime, Pauperism and Heredity. New York: Putnam. Eckersley, R. (1988). Casualties of Change: The Predicament of Young People in Australian: A Report on the Social and Psychological Problems Faced by Young People in Australia. Carlton: Commission for the Future. Eckersley, R. (1992). Apocalypse? No!: Youth and the Challenge to Change. Canberra: Australia's Commission for the Future (Essay Series No. 1). Eckersley, R. (1993). The west's deepening cultural crisis. Futurist, November ± December, pp. 8 ± 12. NP. Foucault, M. (1977). Discipline and Punish. London: Peregrine Books. Gans, H. (1995). The War Against the Poor: The Underclass and Anti-Poverty Policy. New York: Basic Books. Glueck, S., & Glueck, E. (1965). Physique and Delinquency. New York: Harper and Row. Gould, S. J. (1996). Life's Grandeur. London: Jonathan Cape. Hagan, J., & McCarthy, B. (1997). Mean Street: Youth Crime and Homelessness. Cambridge, UK: Cambridge Univ. Press. Haggerty, R., Sherrod, L., Garmezy, N., & Rutter, M. (Eds.). (1994). Stress, Risk and Resilience in Children and Adolescence: Processes, Mechanisms and Interventions. Cambridge, UK: Cambridge Univ. Press. Hawkins, J., Arthur, M. W., & Catalano, R. F. (1995). Preventing substance abuse. In: M. Tonry & D. Farrington (Eds.), Building a Safer Society: Strategic Approaches to Crime Prevention ( pp. 343 ± 427). Chicago: University of Chicago Press. Hawkins, J. D., Herrenkohl, T., Farrington, D. P., Brewer, D. D., Catalano, R. F., & Harachi, T. W. (1998). A review of predictors of youth violence. In: R. Loeber & D. P. Farrington (Eds.), Serious and Violent Juvenile Offenders: Risk Factors and Successful Interventions ( pp. 106 ± 146). Thousand Oaks: Sage Publications. Hawkins, J. D., & Weis, J. G. (1985). The social development model: an integrated approach to delinquency prevention. J Primary Prev 2, 73 ± 97. Herman, A. (1997). The Idea of Decline in Western History. New York: The Free Press. Holloway, W., & Jefferson, T. (1997). The risk society in an age of anxiety: situating fear of crime. Br J Sociol 48, 255 ± 265. Hooton, E. (1939). The American Criminal. Cambridge, MA: Harvard University Press. Jencks, C., & Mayer, S. (1990). The social consequences of growing up in a poor neighbourhood. National Research Council, Inner City Poverty in the United States ( pp. 111 ± 186). Washington, DC: National Academy Press. Langmore, J., & Quiggin, J. (1994). Work for All, Full Employment in the Nineties. Melbourne: Melbourne University Press. Last, J. (1988). A Dictionary of Epidemiology. New York: Oxford Univ. Press. Marvin, W., Sellin, T., & Figlio, R. (1972). Delinquency in a Birth Cohort. Chicago: University of Chicago Press.

J. Bessant / Journal of Criminal Justice 29 (2001) 31±43 McDonald, R. (Ed.). (1997). Youth, the Underclass and Social Exclusion (pp. 1 ± 25). London: Routledge. Neil, C., & Fopp, R. (1992). Homelessness in Australia: Causes and Consequences. CSIRO, Victoria: Division Building, Constructions and Engineering. Nordau, M. (1895, 1968, 1982). Degeneration. Translated from the 2nd ed. (1895). With introduction by Mosse. G. Reprint, New York: Howard Fertig. O'Malley, P. (1992). Risk, power and crime prevention. Econ Soc 21, 252 ± 275. Park, R. (1915). The city: suggestions for the investigation of human behaviour in the city. Am J Sociol 20(5), 577 ± 612. Pearson, G. (1983). Hooligans: A History of Respectable Fears. London: Macmillan. Pick, D. (1993). Faces of Degeneration: A European Disorder 1848 ± 1940. Cambridge: Cambridge Univ. Press. Potas, I., Vining, J., & Wilson, P. (Eds.). (1990). Young People and Crime: Costs and Prevention (pp. 30 ± 42). Canberra: Australian Institute of Criminology. Quiggin, J. (1996). Great Expectations. Sydney: Allen and Unwin. Rehny, N., & McBride, S. (1997). Help Wanted: Economic Security for Youth. Vancouver: Canadian Center for Policy Alternatives. Resnick, M. D., Harris, L. J., & Blum, R. W. (1996). The impact of caring and connectedness on adolescent health and well being. J Pediatr Child Health 29(suppl. 1), 3 ± 9. Rose, N. (1994). Expertise and the government of conduct. Stud Law, Polit Soc 14, 259 ± 367. Rose, N. (1996). The death of the social? Re-figuring the territory of government. Econ Soc 25, 327 ± 356. Rutter, M., Giller, H., & Hagell, A. (1998). Antisocial Behaviour by Young People. Cambridge, UK: Cambridge Univ. Press. Sassan, S. (1991). The Global City. New York: Princeton University Press. Shaw, C., & McKay, D. (1931). Social Factors in Juvenile Delinquency. Washington: USGPO.

43

Shragge, E. (Ed.). (1997). Workfare: Ideology for a New Underclass (pp. 22 ± 29). Toronto: Garamond Press. Sumner, C. (1994). The Sociology of Deviance: An Obituary. Buckingham, UK: Open University Press. Sutherland, E., & Locke, H. (1936). Twenty Thousand Homeless Men: A Study of Unemployed Men in the Chicago Shelters. Philadelphia: JB Lippincott. Sweet, R. (1998). Youth: the rhetoric and the reality of the 1990, Dusseldorp Skills Forum. In: Australia's Youth: Reality and Risk, Dusseldorp Skills Forum, Sydney. ( pp. 2 ± 22). Thomas, W. I. (1923). The Unadjusted Girl: With Cases and Standpoint Analysis. Boston: Little Brown. Thrasher, F. (1927). The Gang: A Study of 1313 Gangs in Chicago. Chicago: University of Chicago Press. Vinson, T. (1999). Unequal in Life: The Distribution of Social Advantage in Victoria and New South Wales. Sydney: The Ignatious Social Policy Center. Watson, I. (1993). The Ideology of the Underclass and the Reality of the Working Poor; Long-Term Unemployment and Occupational Restructuring ( pp. 14 ± 16). University of New South Wales, Sydney: National Social Policy Conference. White, R. (1994). The making of a youth underclass. Youth Stud Aust 13, 19 ± 23. Symposium Paper. Whyte, W. (1943). Street Corner Society. Chicago: University of Chicago Press. Wilson, W. J. (1996). When Disappears: The World of the New Urban Poor. New York: Knopf. Winefield, A., Tiggemann, M., Winefield, H., & Goldney, R. (1993). Growing Up With Unemployment: A Longitudinal Study of its Psychological Impact. London: Routledge. Withers, G., & Batten, M. (1995). Programs for at-risk youth: a review of the American Canadian and British literature since 1984. Dusseldorp Skills Forum, Sydney. Research Monograph No. 47. Wooden, M. (1998). The labour market for young Australians. In: Dusseldorp Skills Forum, Australia's Youth: Reality and Risk, Dusseldorp Skills Forum, Sydney. ( pp. 29 ± 50).