The portrayal of violence in the media: impacts and implications for policy


Numerous research studies identify an association between exposure to violence in entertainment and violent behaviour, but do not prove that exposure causes violent behaviour. Rather, there is a risk that exposure to media violence will increase the likelihood of subsequent aggressive behaviour. This risk can be increased or decreased by a large number of other factors. Appropriate policy responses fall into three general areas: policies that relate to the nature of material that is publicly available, classification issues, scheduling and regulatory enforcement; policies that focus on adult community education and early childhood education that sharpen consumers' ability to discriminate between fantasy and reality, and to become informed and empowered consumers; and policies which focus on health promotion and primary health care, and which treat violence as a public health issue. This Trends and Issues paper seeks to inform the necessary policy debate.

Researchers and professionals have argued for decades about whether or not the portrayal of violence in the various media causes violence in society. Laboratory experiments, field research and correlational studies have all been used as investigative tools. Some studies have suggested that there is a direct causal relationship between violence in entertainment and violent behaviour, and others have concluded that there is no association whatsoever. Most studies have shown that there is some sort of relationship or association.

Most of these studies have focused on television violence and have concluded that there are some negative effects related to watching violent or aggressive behaviour on TV. They do not necessarily indicate a direct cause-and-effect relationship. Rather, they suggest that exposure to media depictions of violence enhances the risk that the viewer will engage in subsequent aggressive behaviour. The effects of exposure to violence in the media are by no means inevitable and may be amplified or reduced by a variety of other factors (Australia 1990). Research into the effects of pornography and violent video/computer games, while less voluminous than television research, has begun to draw similar conclusions.

The relationship between media depictions of violence and subsequent violent behaviour is extremely complex. There are a number of interacting variables which play an important role in determining who will be affected, by what material, and in what way. The context in which the violence is portrayed and age of the viewer/player are the most important variables for determining the potential impact of violence. Also significant is the participant's ability to differentiate between fantasy and reality, and justified or unjustified use of force.

This paper explores the complexities of this problem in the context of the different forms of visual entertainment, and proposes a number of policy options for addressing the issue.

The impact of on-screen violence

The main research findings are:

  • watching violence on screen is related to increased aggression, desensitisation to violence and increased fear of crime;
  • violence in the media may contribute to violent crime, but is not a single cause, because there are many other variables which contribute to violent behaviour;
  • some people may imitate what they see on television and video (and many do not);
  • violence on screen may reinforce the behaviour of already aggressive people;
  • the relationship between viewing violent screen images and exhibiting aggressive behaviour, appears to be bi-directional. That is, aggressive people are more likely to watch violence, and people who watch violence are more likely to be aggressive;
  • the context in which violence is portrayed plays a critically important role in relation to its effects;
  • the effects from on-screen violence can be short or long term;
  • children are most at risk from these effects, and young adults may also be at risk;
  • males appear to be slightly more at risk than females;
  • the general public is concerned about the effects of on-screen violence;
  • parents have an important role to play in supervising their children's viewing, teaching children about the differences between television or film violence and real-life situations, and encouraging critical evaluation of on-screen images;
  • despite the potential influence of violent entertainment on violence in society, it is not clear whether the impact is significant in comparison to the impact of other environmental variables such as family circumstances, violence or abuse in the home, parental influence, poverty, health, education, racism, cultural disintegration, and substance abuse.

There has been criticism of the research linking exposure to media violence and subsequent aggressive behaviour. Freedman (1994) has examined most of the studies (all studies up to 1994) which have been undertaken on this subject and he asserts that there is no evidence in the research to suggest that watching violent TV causes aggression. His argument is largely based on examining the research methods and experimental conditions, recognising that many intervening variables are not taken into consideration or cannot be easily controlled. (Similar criticisms have been made about research which examines the effects of pornography and video/computer games.)

Despite recognising the shortcomings of some types of research, Paik and Comstock (1994) concluded from their meta-analysis of 217 studies, that there was sufficient evidence to suggest an association between watching violence (and erotica) on television and subsequent aggressive behaviour.

Of course it is not contended that television violence is the only cause of aggression and violence in society today. Aggression is a multiply determined behaviour. It is the product of a number of interacting factors-genetic, perinatal, physiological, neurological, environmental. It is only when there is a convergence of factors that violent behaviour occurs.

Prof. Leonard Eron (1995, p. 85)

The importance of context

The effects of watching violence are influenced by the ability of individuals to discriminate between fantasy and reality, between justified and unjustified violence, and the capacity to critically evaluate the portrayal of violence within a social and moral framework.

Research has shown that there are a number of ways that on-screen violence can be portrayed which might influence viewers (see Comstock & Paik 1991; Murray 1994). These include:

  • perpetrators of violent acts who are rewarded or remain unpunished for their actions. To the vulnerable viewer, whether it be a child learning about the world or an already aggressive person, this could be interpreted to mean that violent behaviour is acceptable or even desirable;
  • the aggressive action is seen by the viewer to be justified;
  • there are few or no consequences portrayed, such as remorse, real emotional and physical pain, or legal consequences. The vulnerable viewer who is presented with a false impression of the reality of violence may be more likely to engage in violent behaviour, being unaware of the real consequences;
  • the viewer identifies strongly with the perpetrator, or associates the cues for the violent behaviour with real-life cues, hence the violence appears to be more real. This can reinforce or justify the viewer's own behaviour;
  • the viewer is predisposed to aggression. The violent images can act as a trigger to release these existing feelings;
  • the violence leaves the viewer feeling aroused;
  • something or someone causes the viewer to be frustrated after watching violence;
  • the on-screen violence actually pleases the viewer;
  • the perpetrator's motivation is to cause harm or injury;
  • there is a lack of critical commentary or non-violent balancing events.

The media


Most of the research in this area has examined the effects of television violence. Television is a unique form of entertainment in that almost every home in the "western" world has at least one television. Australians on average watch around 21 hours per week. Because television is a large part of everyday life for many people, images of violence could potentially affect a large proportion of the population, compared to videos and computer games which, while they are increasing in popularity, are watched or played by a smaller proportion of the population. Television programming aims to satisfy the viewing requirements of all age-groups and all types of people and therefore broadcasts a wide range of material, whereas videos and computer games are individually classified and generally chosen according to that classification.

The size of the relation [between TV violence and aggression] is about the same as that between smoking and lung cancer. Not everyone who smokes gets lung cancer and not everyone who has lung cancer smoked. But no one outside the tobacco industry would deny that smoking causes lung cancer. Similarly, not everyone who watches violent TV becomes aggressive and not everyone who is aggressive watched television. But that doesn't mean TV violence is not a cause of aggression.

Prof. Leonard Eron of the American Psychological Association in testimony to the US Senate (1995)

News reporting

As with other forms of violence on television, there is concern that the reporting of suicides, homicides and violent and criminal events on the news, will result in imitation or "Copy Cat" action by some individuals. For example, some studies show that the number of suicide deaths recorded have increased after news reports of suicide (Hassan 1996), particularly where the reports detailed the method used by the victim. Many aspects of the mass murder in Hungerford in England in 1987 were remarkably similar to the Clifton Hill massacre in Australia, and it is thought that the reporting of the Australian tragedy may have triggered the events which took place in Hungerford ten days later (Cantor & Sheehan in press). Similar "Copy Cat" hypotheses have been proposed linking the Port Arthur massacre in Tasmania with the mass murder in Dunblane in Scotland just weeks before.


Videos pose a greater problem than television due to the more explicit, violent material that they can contain within the regulatory guidelines and the fact that such material can be relatively easily accessed. Videos are available almost 24 hours per day and their content ranges from wholesome "feel-good" stories for the whole family, to "X" and "R" rated, violent and sexually explicit. Despite the classification system used, these films are readily available to persons in almost any age-group at any time.

The "Copy Cat" phenomenon as it applies to videos has been used by some people to explain such horrific events as, for example, the killing of a toddler in England by two ten-year-olds, who had apparently watched the movie Child's Play. Although such films might have an influence on the behaviour of some people, it must be emphasised that other factors also contribute to violent behaviour, and not all people who watch violent images exhibit violent behaviour as a result. Most children have seen films intended for age-groups older than their own, and they want to see them because they are naturally curious. Children and adolescents are often unsupervised in their own homes or in friends' homes and can get access to these films. Research has found that the children of parents who watch violent videos are more likely than other children to watch violent videos. It is difficult for parents or legislators to regulate access to videos, but parents in particular must be made aware of the possibility that their children may have access to this material.


Research findings suggest that pornographic films, especially those containing violence, can contribute to callous sexual behaviour and violence towards women (Donnerstein & Linz 1994). This can be due to the manner in which women are portrayed in these films often either as promiscuous or submissive. The images portrayed of women as objects to be used by men for sex, are potentially damaging in the social messages which they present. In their meta-analysis of the available data, Paik and Comstock (1994) found that erotica had a potentially negative effect on the attitude of men towards women, even without incorporating violence. However, as with other research findings about the effect of viewing violent or erotic material, it is important to remember that there are other variables which influence the behaviour of any particular viewer. The context in which this material is portrayed is important, as is the psychological state of the viewer and whether or not they are predisposed to behaving aggressively. The same complex issues which surround the debate about the effects of violence also apply to the effects of pornography.

Video/computer games and beyond

Australian research is currently being undertaken by the Office of Film and Literature Classification into both players' and non-players' perceptions of the portrayal of violence in video/computer games. Many video/ computer games contain some level of violence (although most of these contain "low level" violence), often requiring the player to fight other characters in some sort of battle, injuring or killing them in order to proceed to the next level of the game. Some of the more explicit games have sexual themes, and some have sexually violent themes. The concern expressed about such games is due to the level of interactivity involved. It is as if the player is actively participating in the violence.

As with television or video violence, there is increasing concern that players will become more aggressive as a result of violent games, or will become desensitised to the violence. There is, as yet, little evidence to suggest that a direct causal relationship exists between playing violent video/computer games and becoming violent in real-life. A Senate Select Committee on Community Standards (1993) reported the views of teenage users of these games, who believed that although younger children might be affected by the games, there was less effect as people got older because they could then differentiate between reality and fantasy, and the games were found to be boring after a while.

From 1994 to 1996, the Office of Film and Literature Classification classified 19 per cent of video/computer games as being suitable only for those over the age of 15 years, 1 per cent were refused classification and the remaining 80 per cent were classified as suitable for those under 15 years (29 per cent over 8 years of age, 51 per cent any age).

Although there is now a classification system for these games, the real concern with this type of electronic entertainment is that this material can be easily accessed and downloaded through the Internet. As the Internet cannot be effectively regulated at present, this is extremely difficult if not impossible to avoid, and therefore causes a great deal of concern for parents. There is some software (for example, "Net Nanny") which can be used to limit access to some parts of the Internet but some parents may find this difficult, particularly as the children may be more multimedia information-technology literate than they are. As the pace of technological development increases, the potential for future games is that they will be extremely realistic, as virtual reality has already demonstrated. These games may give players the sensation of actually injuring or killing people. It remains to be seen whether this would lead to an increase or decrease in violent behaviour or homicide rates.

Children in Canada and the United States watch virtually the same television. Yet, the murder rate in Canada, and the rate of violence in general, is much lower than in the United States. Children in Japan watch probably the most violent, the most lurid and graphic television in the world, and the rate of violent crime there is minuscule compared to Canada and the United States.

Jonathan Freedman (1994, p. 854)

Public opinion

Recent thinking by a number of commentators, is that the focus in this debate should be diverted from the question of causality to the question of public opinion. The public is beginning to recognise the potential risk that the portrayal of violence in the media poses, especially for children. The shift appears as psychologists, media industry professionals and other researchers argue about causality, while policy makers search for some evidence which will form the basis for sound policy decisions. The consideration then for policy formulators would be how much to reflect public opinion, the wishes, unscientific commonsense and moral views expressed by ordinary people in society.

Notwithstanding the market for violence in the entertainment media, Gerbner (1994) tells us that in the United States, the ten highest rating shows on television are generally non-violent, and in a study of over 200 shows, half of which were violent and half non-violent, the non-violent shows have had the higher ratings for the past five years. He also tells us that up to 85 per cent of respondents in public opinion polls say there is too much violence on television (Gerbner 1994).

In a recent Morgan poll in Australia, 85 per cent of respondents thought that the media concentrated too much on sensationalist reporting of sex and violence, while 80 per cent thought that the media did not care about people's feelings and rights to privacy. Forty-two per cent thought that television ought to be censored in some way (Murphy 1996).

A study conducted by the Australian Broadcasting Tribunal (ABT) (1990), into TV violence in Australia, documents countless submissions received from the public and worried parents about television news coverage (in particular), and the effect that it might have on children. A number of submissions suggested that the news be broadcast at 9 pm or later (as it is in Britain), instead of the earlier prime-time slot that it currently occupies. The study also found that 60 per cent of adults thought there was too much violence on television, and 72 per cent thought that TV violence would have an effect on children (ABT 1990).

There is some evidence to suggest that the level of violence in television programs, films, newscasts and crime programs may have led to an increase in fear that society is more dangerous than is actually the case. The viewer who identifies strongly with the victim can have increased fear about being attacked, which reinforces the view that society is dangerous for that person (ABT 1990; Sege & Dietz 1994).

Nobody is going to argue about research any more. We in the television industry have acknowledged that we have an influence. We have an impact. We have not said we are blameless in this.

Beth Bressan (1994, p. 859)

Policy implications

Acknowledging the potential impact that watching violent scenes on television or video has on individuals and society, yet noting that different people will respond differently to similar stimuli, the policy challenge is twofold: first, to enhance media violence awareness and responsibility amongst the public; and second, to avoid mechanisms which could trigger unacceptable behaviour in susceptible individuals.

This is best achieved by inter-sectional collaboration, balancing censorship and regulation which have limitations in the extent to which they can control what people see and do; by new technology which can empower individuals to control what information is received or accessed in their own homes; by public health and health promotion campaigns which inform people of the potential harmful effects; and by education campaigns to teach children (and adults) the discriminatory skills necessary for healthy use of the entertainment and information media.

Censorship is generally not a satisfactory option when it involves limiting access to information which a substantial proportion of the public claims it has a right to access. Censorship may also lead to the "forbidden fruit" phenomenon and the development of black markets.

Regulation of broadcasters has similar limitations but self-regulation in response to public demand and opinion, could go some way towards reducing the potential effects, if it can be practised in a truly effective manner. Australian classification and advisory systems are already informing viewers about the nature of program content, and together with responsible scheduling, can help lessen the impact of on-screen violence. One positive action which broadcasters should consider is to ensure that there is a balance of "pro-social" programs showing healthy cooperative behaviour, counteracting the harmful, anti-social messages of violent, aggressive images.

Most of the committees which have investigated this issue have recommended that public health education campaigns, directed particularly at parents, are necessary in order to inform people about the potential detrimental impact that television and video violence can have, especially on children. It is well known that parental influence can be a major factor in reducing the impact that TV/video violence will have on children.

Adult Community Education in Australia reaches over one million people, and crime prevention and violence prevention material could be developed as materials for study circles, to form the basis of awareness discussions. Study circles exist right across the country, and by using them, material could be brought to the attention of many people who would not normally be exposed to crime prevention programs.

Media literacy education for children in all primary and secondary schools is necessary in order to teach children how to critically evaluate the images which are presented to them on an everyday basis, in all forms of electronic entertainment, including video and computer games and the Internet. This could take a similar format to traditional literature classes, or drug and sex education classes, as it is another part of modern life which can be potentially damaging if approached with a lack of relevant knowledge. Education such as this is already occurring in Canada, and the lesson plans are even available on the Internet (YTV Canada Inc. 1994).

The new technologies, which are developing rapidly, are constantly stretching the limits of our imagination and the types of material which we can access. Technological advances such as the "v-chip" (violence chip) can be programmed to block out unwanted programs from television broadcasts. Blocking and filtering software such as "Net Nanny" and "Surf Watch" can similarly be used to limit Internet access. It remains to be determined whether such blocking mechanisms are used responsibly. As with censorship, new technology relies on parental involvement. It requires parents to be aware of what their children are watching or gaining access to.

Six years on, the recommendations made by the Australian Broadcasting Tribunal in 1990 are still relevant, and if fully implemented could have a substantial impact on the effects of media violence. The recommendations highlighted the necessity of an industry-wide code on the treatment of violence on television, with particular care in regard to the child audience, the depiction of suicide and the depiction of violence out of context. It was recommended that the code pay attention to the portrayal of violence, suicide and private grief in news and current affairs programs, and that there be a prohibition on the use of violent acts or scenes in advertising. Also recommended was an awareness campaign to inform the public about violence on television, a nationwide education program to enable parents and schools to help children to deal with violence in television programs, and the Australian Education Council should consider introducing television studies into all primary and secondary schools (Australian Broadcasting Tribunal 1990, vol.1, pp.xxi-xxii).

The policy options outlined above are unlikely to produce results in the short term. More research in a similar vein to that already conducted is not required, but monitoring programs (like the Australian Institute of Criminology Homicide Monitoring Program and its Violence Against Women Indicators Project) could examine the feasibility of including data on the violent viewing habits of offenders.


  • Australia. National Committee on Violence 1990, Violence: Directions for Australia, Australian Institute of Criminology, Canberra.
  • Bressan, B. 1994, "Television networks: Taking affirmative action", Hofstra Law Review, vol. 22, pp. 857-62.
  • Australian Broadcasting Tribunal (ABT) 1990, TV Violence in Australia, Report to the Minister for Transport and Communications, ABT, Sydney.
  • Cantor, C. & Sheehan, P. (in press), "Violence and media reports: A connection with Hungerford?", Archives of Suicide Research.
  • Comstock, G. & Paik, H. 1991, Television and the American Child, Academic Press, San Diego.
  • Donnerstein, E. & Linz, D. 1994, "Sexual violence in the mass media", in Violence and the Law eds. M. Costanzo & S. Oskamp, Sage, United States.
  • Eron, L. 1995, "Media violence", Pediatric Annals, vol. 24, no. 2, pp. 84-7.
  • Freedman, J. 1994, "Viewing television violence does not make people more aggressive", Hofstra Law Review, vol. 22, pp. 833-54.
  • Gerbner, G. 1994, "There is no free market in television", Hofstra Law Review, vol. 22, pp. 879-84.
  • Hassan, R. 1996, Social Factors in Suicide in Australia, Trends and Issues No. 52, Australian Institute of Criminology, Canberra.
  • Murphy, D. 1996, "Facts and friction", The Bulletin, 28 May, p.14.
  • Murray, J. 1994, "The impact of televised violence", Hofstra Law Review, vol. 22, pp. 809-25.
  • Paik, H. & Comstock, G. 1994, "The effects of television violence on antisocial behavior: A meta-analysis", Communication Research, vol. 21, no. 4, pp. 516-46.
  • Sege, R. & Dietz, W. 1994, "Television viewing and violence in children: The pediatrician as agent for change", Pediatrics, vol. 94, no. 4, pp. 600-7.
  • Senate Select Committee on Community Standards Relevant to the Supply of Services Utilising Electronic Technologies 1993, Report on Video and Computer Games and Classification Issues, Parliament of the Commonwealth of Australia, Canberra.
  • YTV Canada Inc. 1994, "Television and Violence Lesson Plan"


The Australian Institute of Criminology acknowledges the assistance provided in the preparation of this paper by Professor Peter Sheehan, Professor of Psychology, University of Queensland.