OUTSIDE DISCOS

HOT SPOTS FOR VIOLENCE: THE ENVIRONMENT OF PUBS AND CLUBS


In Villa Gessel, a   well known seaside resort  south of Buenos Aires, a young descendant of a paraguayan family was beaten to death outside a nearby disco by a group of   thugs who practised rugby
 Did the homicide have to do with Rugby ethos , with drugs and alcohol abuse or with other deeply rooted factors of immorality in our youths ?



What is it like on first world countries like Australia ? 


Professor Ross Homel School of Justice Administration Griffith University Queensland Dr Steve Tomsen Department of Sociology University of Newcastle New South Wales

VIOLENCE HAS BECOME A MATTER OF MAJOR PUBLIC CONCERN IN AUSTRALIA in recent years.


Two separate massacres in Melbourne in 1987 and one in Sydney in 1991— each by a lone gunman— horrified the nation and reinforced the view of many that Australia is becoming a more violent and lawless country. Anecdotal evidence and systematic survey data both confirm that fear of random, unprovoked violence from strangers now has a major effect on the lives of many ordinary Australians (van Dijk, Mayhew & Killias 1991). However, criminological research (see Australia 1990) suggests that most acts of interpersonal violence are not the result of random attacks by madmen on complete strangers, but involve ordinary people as both attackers and victims who frequently know each other and who, for one reason or another, come into conflict in the home, street, workplace, or place of recreation. Indeed, even a quick perusal of the statistics on homicides and assaults leads one to a conclusion which may be banal but is of fundamental importance: the places in which most acts of interpersonal violence occur, and the times at which they occur, mirror, at least roughly, the rhythms and routines of daily Homicide: Patterns, Prevention and Control 54 life (see Robb 1988). A substantial number of homicides involve intimates within the family home because those are the circumstances in which many people spend much of their lives. A higher proportion of non-fatal assaults than of homicides take place outside the home and involve strangers, but these events are also highly patterned, being more common late at night and on weekends, when social life and interaction is at its most intense. As Cohen and Felson put it: Rather than assuming that predatory crime is simply an indicator of social breakdown, one might take it as a byproduct of freedom and prosperity as they manifest themselves in the routine activities of everyday life. (1979, p. 605). There are few activities as routine in Australian culture as the imbibing of alcoholic beverages. According to a 1988 survey of four Australian states (Berger et al. 1990), three-quarters of all adults drink at least occasionally and one in ten can be classified as a 'heavy drinker'. Half the population drink at least once or twice a week, and especially for men under the age of twenty-five, this drinking is often done at hotels or licensed clubs. It is perhaps not surprising, therefore, that assaults and homicides frequently involve the presence of alcohol in the offender, victim, or both. Collins (1989) cites evidence that as many as 80 per cent of those arrested in the USA for cutting, concealed weapons, other assaults, murder, and shooting had measurable levels of alcohol in their urine, while Robb (1988)— in a study of serious assaults recorded by police in New South Wales— found that 40 per cent were nominated by police as involving alcohol. Moreover, assaults coming to police notice and recorded by them frequently occur after midnight around pub closing times, and at least 20 per cent take place in or around licensed premises (Victoria 1989; Robb 1988). Alcohol involvement in homicides occurring during peak entertainment periods seems to be even more pronounced than for non-fatal assaults: nearly half (48 per cent) of all homicides occurring on Saturdays in New South Wales in the years 1968 to 1986 involved suspects who had been drinking in the previous twelve hours (Bonney 1987). However, police statistics may greatly understate the extent of alcohol-related violent crime because most assaults are not reported and because police often seem to be more reluctant to record reported alcohol-related assaults than reported nonalcohol-related assaults. Surveys of injured persons presenting at hospital for treatment permit a clearer picture of the true incidence of pub and club related violence, since assault victims are more likely to seek medical than police assistance. One recent hospital survey in Sydney suggests that each year in New South Wales many thousands of people, mostly young men, are injured (sometimes quite seriously) as a result of assaults occurring in or around licensed premises (Cuthbert 1990). Statistical associations do not prove that alcohol consumption actually causes violence. After all, if drinking is common behaviour, and if violent incidents can be thought of as 'routine activities which share many attributes of, and are interdependent with, other routine activities' (Cohen & Felson 1979, p. 589), then it would be surprising not to find alcohol implicated in many instances of assault. The question is whether alcohol consumption itself Hot Spots for Violence: the Environment of Pubs and Clubs 55 contributes in some way to the likelihood of violence, or whether aspects of the drinkers or of the drinking settings are the critical factors. It is quite possible, for example, that male attitudes which legitimise the physical maltreatment of women, or environmental factors like crowding, discomfort, and aggressive bouncers in pubs and clubs, are the real causes of much alcohol-related violence (McGregor 1990; Victorian Community Council Against Violence 1990).


The purpose of this paper is to report briefly the method, results, and implications of some observational research into pubs and clubs which were conducted in Sydney in 1989 (for further details of the study see Tomsen, Homel & Thommeny 1990; Homel & Tomsen 1991; Homel, Tomsen & Thommeny 1992). The study was the first systematic attempt in Australia to examine possible links between aspects of the environment of public drinking and the occurrence of violence. A key assumption was that there is a complex (but nevertheless real) relation between violence and public drinking (not the mere ingestion of ethanol) which is embedded in Australian history and culture and reproduced in institutional arrangements and regulatory and police practices regarding drinking. This research aimed to transcend the narrow debate about the effects of ethanol the substance by focusing on the total environment of drinking and its regulation (or lack of regulation) by management, police, and other public officials. Thus features of the external regulation of licensed premises were considered as well as more directly observable characteristics such as physical layout, patron mix, and social atmosphere. Although not directly a study of homicide, the research has obvious relevance if one accepts the assumption that violence occurring in and around licensed premises forms a continuum, from acts of non-physical aggression through 'brawls' and 'fights' to acts of homicide. While obviously most incidents are relatively minor, occasionally people are killed in or around clubs and pubs. Indeed, more than one male homicide victim in ten is killed in these locations (although, interestingly, fewer than 2 per cent of female victims: Bonney 1987).


On this view, a homicide is a serious assault which for some reason results in death rather than serious injury, but for which the situational factors are not qualitatively different from those applying to non-fatal incidents. This assumption is open to question, and should be empirically investigated. However, since no homicides were observed in the course of this study, little light can be thrown on the issue in the present paper, except to point out that some of the worst premises in Sydney were included in our observations.

Method Studies of drinking in public places have been conducted for many years (Fisher 1985). A study by Graham and her colleagues (Graham et al. 1980) in Vancouver was especially valuable as a guide for this research, since data were obtained for a large number of situational variables as well as for instances of aggression and physical violence. Four observers (working in male-female pairs) noted 160 incidents of aggression (forty-seven involving physical violence) in 633 hours of observation in 185 drinking establishments.

Many variables were positively Homicide: Patterns, Prevention and Control 56 correlated with aggression, including the percentage of drunk patrons, the percentage of American Indians, poor ventilation, the amount of sexual body contact, lack of cleanliness, and a hostile atmosphere. The authors stressed, however, that the barroom environment is best viewed as 'an ecological system', and implied that the overall influence of this ecology on aggression may be greater than the sum of the effects of individual variables. Through factor analysis and qualitative analysis they identified one distinct type of bar— the Skid Row Aggressive Bar— which was characterised by high levels of aggression, extreme intoxication, down-and-out patrons not accepted anywhere else, and a 'bizarre atmosphere' in which deviant and unusual behaviours were tolerated, and even encouraged. The extent to which skid row bars constitute a major part of the problem of violence in other localities is a matter for further research. Many of the variables and insights from the Vancouver research were used as a starting point for this study. Influence also came from these authors' suggestions that future research concentrate on places where alcohol-related aggression most often occurs, and that within this context more details be collected on the processes of aggression. This immediately raised two related questions: the method of sampling (how were 'high risk' venues to be identified?); and the method of data collection (how best could 'ecological processes' be studied?).


An early decision was taken to use qualitative rather than quantitative methods, relying heavily on unstructured observations in licensed premises and, to a lesser extent, on semi-structured interviews with licensing and general duties police, chamber magistrates, and security industry personnel. There were several reasons for the decision to use a qualitative approach. While the Vancouver research identified many potentially relevant situational factors, it was obvious reading the results that there were a large number of differences between drinking settings in Sydney in the late 1980s and in Vancouver in the late 1970s, and that many more variables would have to be generated. More importantly, it was judged that the reduction of the problem to 'variable analysis', even after extensive piloting, would hinder attempts to explore the complex interactions and subtle processes which, we hypothesised, led to violence. In addition, the absence of any objective database in New South Wales identifying licensed premises as more or less violent necessitated a 'theoretical sampling' strategy based on the best available qualitative judgements concerning premises' standings in terms of violence and/or poor management (Glaser & Strauss 1967). Being unable to stratify the population of premises, unweighted probability sampling would have yielded too few high risk premises to allow detailed analysis of the relevant environmental factors. The aim was to contrast situational variables and management practices in a small number of premises known to have been regularly violent over a long period of time with the same factors in a sample of establishments noted for their lack of violence or for their ability to defuse violent incidents when they occurred. Using this design, even if little violence was actually observed in the study, it would be possible to explore aspects of drinking settings which were associated with violence. Eventually four highrisk and two low-risk premises were identified on the basis of first-hand knowledge, police Hot Spots for Violence: the Environment of Pubs and Clubs 57 information and exploratory visits. Each of these premises was visited by pairs of observers at least five times, each observation period being between two and six hours in duration. A further sixteen sites were visited at least once, making a total of fiftyfive visits to twenty-three sites in seventeen establishments. Total observation time was nearly 300 hours. All the premises studied intensively were in suburban locations. This was not because it was thought that there is more drinking-related violence in the suburbs— police statistics and research suggest the contrary— but because the problems in city locations such as Kings Cross in Sydney or West End in Melbourne are often dispersed across a number of violent venues. It is easier to plausibly link public violence and the characteristics of particular premises in suburban locations. One consequence of this sampling strategy was that 'skid row' premises did not figure as prominently as in the Vancouver study, although some were visited, particularly during the exploratory phases. One type of problem location studied intensively was licensed clubs. Licensed clubs have often been credited with being more orderly than hotels and having good control over their patrons, but it appears that financial pressures have led many to develop forms of entertainment, principally late-night discos or live music for young people, which create problems not anticipated by management. In fact, all the violent mainstream premises chosen for full study, whether in pubs or clubs, traded after midnight and were popular with young people because they provided live music or a disco. The control sites were studied for the features which distinguish them from the violent. However, during field research it soon became apparent that the violent premises are for most of the time not violent. Violent occasions in these places seemed to have characteristics that clearly marked them out from non-violent times. In effect, these locations were acting as controls for themselves. This unexpectedly helped to refine ideas about the relevant situational variables and to some extent reduced the importance of comparisons with the premises selected as controls. Eventually it could be noted which variables regularly prevailed and linked up with each other during violent or peaceful periods. This method was actually an extension of the theoretical sampling approach, from the choice of physical sites to the choice of times and days for observation. For example, one site was very violent on a night which had discount drinks and a punk band. The observers returned on a similar night with full-priced drinks and a similar band so as to discern any differences and hoping that other aspects of the drinking environment were essentially unchanged. By this, comparative means hypotheses about key variables and their relations were being tested out during observation and to some extent were directing its pattern. In broad terms, the method of data collection and analysis was based on Miles and Huberman (1984). After leaving a site, notes were made and information written onto an observation sheet listing a large number of variables. The visits were then written up as separate narrative accounts by each observer. These narratives were cross-checked and later coded at group meetings. Early and often clumsy efforts to code narratives signalled whether or not various codes were meaningful and linked to the data. Indeed, the real Homicide: Patterns, Prevention and Control 58 thinking about the research questions took the form of regularly revising the codes, a process which was more productive than premature speculation about the 'real' causes of violence. After at least a dozen revisions, almost 200 coded items were settled upon, grouped under a small number of broad headings: physical and social atmosphere, drinking, patrons, staff, and violence. To simplify analysis, summary sheets for each narrative were also prepared, classified according to the degree of violence observed. Results In total, thirty-two assaults involving some degree of physical violence were observed during the course of the study. Excluding nine rough ejections which were borderline assaults, the thirty-two incidents represent a rate of about 11 per 100 hours observation— 50 per cent higher than the rate of 7.4 in the Vancouver study. The higher rate is not surprising, since violent premises were over-sampled and were also sampled late at night when violence is more likely. Graham et al. (1980) report that they witnessed no brawls and no incidents involving serious physical injury.





By contrast, four of the thirty-two incidents in our study were 'brawls', and a number of assaults were rated as 'serious' by observers. For example, in one case an ejected patron was held by three bouncers who repeatedly bashed his head against a steel garbage crate, while on another occasion a floor manager apparently confused a young man with another patron who had been in a fight, lost his temper, and began to throttle this very small and young drinker who choked for several minutes. In contrast to findings from the limited amount of anthropological research, such as Dyck's (1980) analysis of 'scrapping' in barrooms in the small western Canadian city of Parklund, most of the incidents in our study could not be characterised as 'fights' in the sense of being equal conflicts freely entered into by the participants. In at most one-quarter of cases could the victim be said to have actually or possibly invited the attack. Assailants— whether patrons or staff— who deliberately seek out a violent encounter appear to pick their mark, who are most often fewer in number, younger, and smaller. Assailants also appear to focus on victims who they see are quite drunk, or at least far more intoxicated than they are. When more than two parties are involved, pub assaults are often further trivialised as 'brawls', with the equal responsibility of all parties— assailants and victims— implied by this. By our reckoning, equal responsibility is usually not the case. The data analysis suggests that much of the violence observed in this study was not due to anything inherent in public drinking or in the typical patrons of these venues. The key variables suggested by the 'constant comparative' and empirically grounded form of analysis were aspects of the patron type, the social atmosphere, drinking patterns, and the behaviour of doormen. Patron type The typical patrons in violent premises are young, working class men. However, the social class of patrons cannot explain the differences between these violent sites and more peaceful venues with patrons from a similar social background, nor can it explain why the violent venues are at other Hot Spots for Violence: the Environment of Pubs and Clubs 59 times peaceful, although the patrons present are much the same. Moreover, there is no clear causal connection between the young age of regular patrons and levels of violence. The single venue with the greatest number of young drinkers was in fact one of the non-violent locations selected as a control. The gender ratio in venues, together with the social links between the males present, seem more critical than age or class. The proportion of males and presence of male groups in any venue seem to exacerbate feelings of rivalry and group loyalty and can result in arguments and fights. Males in groups, especially as strangers to each other, were seen to come into conflict more readily. The venues we studied drew a larger number of these groups of strangers than others, attracting people from a fairly wide area. Solo males and males with female partners or in mixed groups appear to be less inclined to enter into conflicts. However, it should not be assumed that the presence of women always has a pacifying effect on social atmosphere or that it is rare for women to become involved in conflict and violence. Women were victims of male violence in two incidents, and arguments, challenges, mock fights and fights between females were observed regularly at most of the violent locations— the worst of these involving women bouncers. A surprisingly high number of women patrons spoken to also followed the male path in being apparently indifferent to acts of violence, and several took pleasure in watching fights and brawls. Social atmosphere Although no direct connection between physical attractiveness and violence could be found, attractive, renovated, and well-designed surroundings commonly mean that a venue also has a responsible management and positive staff who relate well to patrons. Not surprisingly, unattractive, neglected, and dirty venues also tended to be among the least comfortable and to have poorly supervised, aggressive, and abrasive staff. However, despite all the myths, rough pubs with plenty of rowdy behaviour (which would include the local workingmen's pubs celebrated in Australian folklore) are not necessarily violent. The two most relevant aspects of atmosphere seem to be comfort and boredom. Comfortable premises are not necessarily the most attractive, renovated places. The most important aspects seem to be roominess, ventilation, and, especially if it is from music of poor quality, only moderate noise. Big crowds in most sites usually mean discomfort for many patrons— a problem exacerbated by a lack of seating and by crowded corridors, stairs, and doorways. Patrons in these situations tend to alleviate their discomfort by more rapid drinking, which causes higher levels of drunkenness, and eventually aggressive reactions to discomfort directed at individuals or property. Overcrowding on dancefloors appeared to be linked to several arguments and at least one of the severe assaults observed. Levels of comfort interact with levels of boredom. Entertained crowds are less hostile, drink more slowly, and seem to be less bothered by uncomfortable surroundings. In some venues, levels of boredom and aggression directed at other patrons were reduced by entertainment in such Homicide: Patterns, Prevention and Control 60 forms as television, videos, and game and card machines. Stage entertainment including dancing and quizzes were also noted to reduce levels of boredom and aggression, sometimes at a critical point during the night when the form of patron interaction suggested that conflict was likely. Of the many aspects of entertainment and boredom, bands and music are perhaps of greatest importance.
While violent and non-violent occasions do not follow a simple bands/no bands dichotomy, quality bands that entertain an audience generate a positive social atmosphere that has been observed to counteract other negative variables. A smaller crowd with a bad band seems more likely to present trouble than a large crowd entertained by quality musicians. Drinking patterns High levels of intoxication are an obvious feature of many violent occasions. This is worsened by discount drinks, with prices in some venues being as low as ten or eleven cents. More commonly, discount drinks are set at around a dollar on specific discount nights. On these occasions many patrons who have paid a high cover charge (for example, $10) in order to see an inferior band or just to enter a disco, seem to decide that they should become quite drunk in order to get their 'money's worth'. In fact, the most violent visit of all, with very high levels of intoxication and seven assaults observed in a few hours, was an 11c discount night with an $11 cover charge. In some clubs, cheaper drink prices (for example, $1.30 for a strong mixed drink) can also serve to bring on very high levels of drunkenness and resulting violence. This was obviously the case in the licensed club included among our group of four most violent premises. As already noted, it is important that these rates of drinking can also be artificially raised by high discomfort and boredom. Many patrons appear to pass through stages of drunkenness— with aggression coming later. Substantial amounts of food that can lower levels of drunkenness were generally not available in the violent premises we studied, especially later at night when patrons are more intoxicated. Some locations had a small range of hot food available, but more often snacks were limited to hotdogs. Hotdog stands often appear to have the adverse effect of encouraging patrons to mill around outside venues. Both of the non-violent control locations that were studied operated substantial restaurants. Doormen The behaviour of bar staff does not figure as highly as expected in the creation of an aggressive or violent atmosphere. Edgy and aggressive bouncers are another matter. They have been observed to initiate fights or further encourage them on several occasions. Some were even observed to leave premises while they were on duty in order to continue a fight with departing patrons. More often, they have been seen to show a good measure of indifference to violence. They regularly asked conflicting patrons to leave premises, and then virtually arranged a fight that they and departing patrons could watch immediately outside the location. The unprofessional view that Hot Spots for Violence: the Environment of Pubs and Clubs 61 assaults occurring just outside the premises where they are employed are not their business, appears to be commonplace. Many bouncers seem poorly trained, obsessed with their own machismo (relating badly to groups of male strangers), and some of them appear to regard their employment as giving them a licence to assault people. This may be encouraged by management adherence to a repressive model of supervision of patrons ('if they play up, thump 'em') which, despite their belief, does not reduce trouble and adds further to a hostile and aggressive atmosphere. In practice many bouncers are not well-managed in their work and appear to be given a job autonomy and discretion that they cannot handle well. The bad relations of many male bouncers with male patrons led some of our informants to suggest that women should be employed on pub and club doors. Although in one club the use of a well-spoken female on the door seemed to appease groups of males who were refused entry for non-membership, it is simplistic to suggest that these sorts of conciliatory skills are held by all females and no males.
Some male door staff were observed to have these qualities and took the role of restraining other bouncers from excessive violence. The most relevant factors seem to be training and experience, rather than gender. Unfortunately, bouncing is still an occupation with a high rate of turnover. Younger bouncers may be leaving this sort of work just as they are beginning to acquire the sort of experience and work maturity that their job requires. Summary Violent incidents in public drinking locations do not occur simply because of the presence of young or rough patrons or because of rock bands, or any other single variable. Violent occasions are characterised by subtle interactions of several variables. Chief among these are groups of male strangers, low comfort, high boredom, high drunkenness, as well as aggressive and unreasonable bouncers and floor staff.
Discussion This research confirms the work of Graham et al. (1979) and others in that a great deal of violence occurs in and around licensed premises. While some of the violence observed did not result in serious injury, many of the incidents would be classified as serious assaults by any reasonable criterion. In this respect, the research findings reflect everyday experience that violence is a routine aspect of interactions in many pubs and clubs. However, it is important to recall that the sample of Sydney drinking places was biased towards times and places where prior knowledge indicated that violence was likely to occur and that, even in the worst places, many visits were 'uneventful' in the sense that no violence was observed. This is consistent with the finding of Graham and her colleagues (1979) that a small number of premises— chiefly of the 'skid row' variety— accounted for a high proportion of all observed instances of aggression. It is also consistent with the preliminary results of a recent observational study of a representative sample of thirty-four Sydney clubs and pubs which found that 16 per cent of Homicide: Patterns, Prevention and Control 62 all premises accounted for three-quarters of all observed incidents of physical violence. (This study was supervised by Ross Homel and involved 147 visits and 300 hours of observation. It was carried out by twenty-two senior year students of Macquarie University in July and August 1991.) The existence of 'hot spots' of predatory crime has been the subject of some recent criminological research. Sherman, Gartin and Buerger (1989) analysed calls to police in Minneapolis over one year and showed that all recorded domestic disturbances occurred at 9 per cent of all possible addresses, while all recorded assaults occurred in only 7 per cent of all possible locations in the city. These authors build on a popular sociological theory of crime, 'routine activities theory', which attempts to account for the non-random distribution of crime by proposing that the rate at which such events occur in collectivities is affected by: the convergence in space and time of the three minimal elements of direct-contact predatory violations: motivated offenders, suitable targets, and the absence of capable guardians against a violation (Cohen & Felson 1979, p. 589). Sherman and his colleagues extend the ideas of routine activities theory from collectivity to 'place', arguing that places, like persons, can be seen to have routine activities subject to both formal and informal regulation. One advantage of applying the routine activities perspective is that it becomes immediately apparent that single variable theories of violence (for example, 'he did it because he was drunk') are unlikely to have much explanatory power. This is because the critical factors are not those to do with offenders, victims, or guardians alone, but those affecting their convergence in time and space. Thus high rates of intoxication do not on their own guarantee that violence will break out, since it is not clear that intoxication will inevitably increase the supply of motivated offenders and suitable victims, or that it will have any effect on the presence of capable guardians.
However, in interaction with other factors, intoxication may be a potent explanatory factor— as this analysis suggests. Laboratory research has generally failed to find any direct connection between the ingestion of alcohol and the incidence of aggression (see Gustafson 1986a; Taylor & Gammon 1976), but has highlighted the importance of interactions of alcohol consumption with factors like frustration (Gustafson 1986b). Violence may therefore occur (as it did in this study) when some patrons are vulnerable to attack due to their extreme intoxication, when formal or informal controls are not sufficient to deter violence, and when potential offenders are drunk and frustrated— frustration perhaps being promoted by poor quality entertainment or by crowding. In a well-managed club or pub employing skilled doormen and floor staff who can detect problem situations before they get out of hand, or who have good communication skills and can defuse aggression before it leads to violence, there may rarely be any connection between levels of intoxication and violence. Alternatively, aspects of patron mix may amplify or reduce the risk of violence when rates of intoxication are high by affecting the processes of informal guardianship or by influencing the motivations of offenders and Hot Spots for Violence: the Environment of Pubs and Clubs 63 the supply of victims. Drunk males on their own often make good victims; informal controls on aggression may work far better in groups consisting of both men and women, even if everyone is drunk, than in all-male groups. The emphasis on the interactions of several factors does not mean that the need for some direct controls on intoxication is rejected. One striking aspect of poor management is the way in which some licensees promote high levels of drunkenness by various kinds of drinks promotions, such as cheap drinks combined with high cover charges. For this reason, an immediate and direct legislative assault on all practices involving discount drinks, 'twofor-one' promotions, happy hours, and any other serving practices which have the effect of producing high levels of drunkenness in a short period is advocated. While one might debate the rights of individual patrons to choose to drink to intoxication, our findings concerning the destructive effects of mass binge drinking resulting from deliberate and irresponsible price discounting and drinks promotions leaves us in no doubt that such practices should be banned. Other major policy recommendations also concern the responsibilities of management. A major flaw in the current form of the Liquor Act 1982 (NSW) is that violence is mentioned in passing in only two places, with assaults on individual victims being seen as the responsibility of those victims, rather than being viewed as the outcome of management practices. There is an urgent need for amendments to this Act so that the continuous operation of a violent venue is an offence that will lead to the cancellation of a licence. Action to close down at least five regularly violent discos in the West End area of Melbourne has been taken in the past two years, on the initiative of the Liquor Licensing Commission and the Victorian Community Council Against Violence, but similar action appears never to have been contemplated in New South Wales and other states. A further policy priority to emerge from this research was the need for better regulation and training of bouncers. Bouncers are required under the Security Amendment Act 1985 (NSW) to hold both a valid and current security licence and to carry related identification on the job, such as a photo identification card. One aim of this legislative provision is to discourage aggressive and violent individuals from becoming bouncers. However, it is obvious from research that, despite the intentions of the legislators, a significant number of working (and licensed) doormen are still prone to violence. This suggests that there is a need to mandate training for security staff in human interaction skills, crowd control, and non-violent conflict resolution. In addition, it is essential that the existing legislation be enforced in line with the intentions of parliament. Without a greater overall police effort to implement the Security Amendment Act, it has little more than symbolic value and will do nothing to reduce actual levels of violence.




Conclusion Regular violence in public drinking locations cannot simply be blamed on rowdy patrons or excused as something natural and unstoppable. Nor can it simply be blamed on the irresponsible ingestion of a legal drug. The drinking environment is an evolving historical and cultural product which can be left Homicide: Patterns, Prevention and Control 64 unchanged or altered for the better. It is clear from our research that continuous patterns of violence in these locations are strongly related to local situational variables, which in turn reflect management practices and government legislation and regulation. While it is difficult, and perhaps not even desirable, to attempt to modify the routine activities of pub and club goers, it is far easier, and surely consistent with broad considerations of the public good, to regulate the routine activities of the premises they frequent (Sherman et al. 1989). The most extreme way of doing this is to incapacitate the activities of the worst 'hot spots' by licence cancellation. Short of this, the other reforms in regulatory practices which have been outlined briefly in this paper have the potential to improve greatly the safety of licensed premises for the tens of thousands of young people who rely on these places to provide most of their entertainment. Acknowledgments The study of licensed premises in Sydney was funded by the Criminology Research Council and the New South Wales Directorate of the Drug Offensive, and was also actively supported by the New South Wales Police Department, who seconded Constable J. Thommeny to the project to assist in data collection and analysis. The opinions expressed in this paper are those of the authors and do not necessarily reflect the views or policies of the funding agencies or of the New South Wales Police Service. References Australia. National Committee on Violence 1990, Violence: Directions for Australia, Australian Institute of Criminology, Canberra. Berger, D.E., Snortum, J.R., Homel, R.J., Hauge, R., & Loxley, W. 1990, 'Deterrence and prevention of alcohol-impaired driving in Australia, the United States, and Norway', Justice Quarterly, vol. 7, pp. 453–65. Bonney, R. 1987, Homicide 2, New South Wales Bureau of Crime Statistics and Research, Sydney. Cohen, L.E. & Felson, M. 1979, 'Social change and crime rate trends: a routine activity approach', American Sociological Review, vol. 44, pp. 588–608. Collins, J.J. 1989, 'Alcohol and interpersonal violence: less than meets the eye', in Pathways to Criminal Violence, N.A. Weiner & M.E. Wolfgang, SAGE, Newbury Park, pp. 49–67. Cuthbert, M. 1990, 'Investigation of the incidence and analysis of cases of alleged violence reporting to St Vincent's Hospital', in Australian Violence: Contemporary Perspectives, eds D. Chappell, P. Grabosky, & H. Strang, Australian Institute of Criminology, Canberra, pp. 135–46. Hot Spots for Violence: the Environment of Pubs and Clubs 65 Dyck, N. 1980, 'Booze, barrooms and scrapping: masculinity and violence in a Western Canadian town', Canadian Journal of Anthropology, vol. 1, pp. 191–8. Fisher, H.R. 1985, Studies of Drinking in Public Places: An Annotated Bibliography, Addiction Research Foundation, Toronto. Glaser, B.G. & Strauss, A.L. 1967, The Discovery of Grounded Theory: Strategies for Qualitative Research, Aldine, New York. Graham, K., La Rocque, L., Yetman, R., Ross, T.J., & Guistra, E. 1980, 'Aggression and barroom environments', Journal of Studies on Alcohol, vol. 41, pp. 277–92. Gustafson, R. 1986a, 'Threat as a determinant of alcohol-related aggression', Psychological Reports, vol. 56, pp. 287–97. Gustafson, R. 1986b, 'Alcohol, frustration and aggression: an experiment using the balanced placebo design', Psychological Reports, vol. 59, pp. 207–18. Homel, R. & Tomsen, S. 1991, 'Pubs and violence: violence, public drinking and public policy', Current Affairs Bulletin, vol. 68, no. 7, pp. 22–7. Homel, R., Tomsen, S. & Thommeny, J. 1992, 'Public drinking and violence: not just an alcohol problem', Journal of Drug Issues, vol. 22, no. 3, pp. 679–97. McGregor, H. 1990, 'Domestic violence: alcohol and other distractions— a grassroots perspective', in Alcohol and Crime, ed. J. Vernon, Australian Institute of Criminology, Canberra, pp. 59–66. Miles, M. & Huberman, A. 1984, Qualitative Data Analysis, SAGE, Beverly Hills. Robb, T. 1988, Police Reports of Serious Assault in New South Wales, New South Wales Bureau of Crime Statistics and Research, Sydney. Sherman, L.W., Gartin, P.R., & Buerger, M.E. 1989, 'Hot spots of predatory crime: routine activities and the criminology of place', Criminology, vol.