2003 Journal of Criminal Justice and
Journal of Criminal Justice and Popular Culture, 10(2) (2003) 109-126
MEDIA CONSUMPTION AND PUBLIC ATTITUDES TOWARD CRIME AND
JUSTICE: THE RELATIONSHIP BETWEEN FEAR OF CRIME, PUNITIVE ATTITUDES, AND
PERCEIVED POLICE EFFECTIVENESS
Western society is fascinated with crime and justice. From films, books, newspapers, magazines, television broadcasts, to everyday conversations, we are constantly engaging in crime "talk". The mass media play an important role in the construction of criminality and the criminal justice system. The publicís perception of victims, criminals, deviants, and law enforcement officials is largely determined by their portrayal in the mass media. Research indicates that the majority of public knowledge about crime and justice is derived from the media (Roberts and Doob, 1990; Surette, 1998). Therefore, it is imperative to examine the effects that the mass media have on attitudes toward crime and justice. The purpose of this research is to examine how the media influences audience perceptions of police effectiveness and to examine whether media consumption is related to fear of crime and punitive justice attitudes.
FEAR OF CRIME AND PUNITIVE JUSTICE ATTITUDES
Research on the effect that the media has on the public revolves around two interconnected issues. Does coverage of sensationalistic and violent crime create fear among the general public and does this fear influence criminal justice policy attitudes? Review of the [End page 109] research indicates that there are mixed results regarding the influence of the news media on creating an attitude of fear among the general public (Surette, 1998). In an early study, Gerbner et al (1980) hypothesized that heavy viewing of television violence leads to fear rather than aggression. Gerbner et al (1980) find that individuals who watch a large amount of television are more likely to feel a greater threat from crime, believe crime is more prevalent than statistics indicate, and take more precautions against crime. They find that crime portrayed on television is significantly more violent, random, and dangerous than crime in the "real" world. The researchers argue that viewers internalize these images and develop a "mean world view" or a scary image of reality. This view is characterized by "mistrust, cynicism, alienation, and perceptions of higher than average levels of threat of crime in society" (Surette, 1990:8). Further studies on the relationship between fear and television viewing indicate a direct and strong relationship (Barille, 1984; Bryant, Carveth and Brown, 1981; Hawkins and Pingree, 1980; Morgan, 1983; Williams, Zabrack and Joy, 1982, Weaver and Wakshlag, 1986). Conversely, Rice and Anderson (1990) find a weak, positive association between television viewing and fear of crime, alienation and distrust. However, multiple regression analysis fails to support the hypothesis that television viewing has a direct, substantial effect on fear of crime.
In a review of the research, Heath and Gilbert (1996) find that the relationship between media presentations and crime is dependent on characteristics of the message and the audience. Presentation of large amounts of local crime news engenders increased fear among the larger public, (Brillon, 1987; Sheley and Ashkins, 1981) while the presentation of large amounts of non-local crime news has the opposite effect by making the local viewers feel safe in comparison to other areas (Liska and Baccaglini, 1990). In addition Chiricos et al (2000) finds that local and national news are related to fear of crime. The effect of local news on fear of crime is stronger for residents in high crime areas and those who experienced victimization.
In terms of audience effects, fear of victimization will depend on who is viewing the crime stories. Research indicates that residents in high crime urban areas who watch a large amount of television are more likely to be afraid of crime (Doob and MacDonald, 1979; Gerbner et al, 1980). Another important factor is whether audience members have direct victim experience or share characteristics that make them crime vulnerable. Research indicates that media sources will be more meaningful when direct experience is lacking (Gunter, 1987; Liska and Baccaglini, 1990; Skogan and Maxfield, 1981). For example, Liska and Baccaglini (1990) find that media influence was strongest for females, whites and the elderly, which are segments of the population least likely to be victimized. In another study, Chiricos et al (1997) find that the frequency of watching television news and listening to the news on the radio is significantly related to fear. Their research indicates that television news consumption is significantly related to fear only for white females between the ages of 30 and 44. This is similar to other findings that suggest that watching crime on television has a greater effect for women and whites, who have low victim risk compared to males and non whites (Gerbner et al., 1980).
Examining the National Opinion on Crime and Justice (NOCJS), Haghighi and Sorensen (1996) find that local media attention to crime was significantly related to fear of sexual assault; getting mugged, beaten up, knifed or shot; and being burglarized while at home. Fear of crime was not significantly related to fear of car-jacking, being murdered, or being burglarized while not at home. However, they did not find the source of crime news to be a factor in fear of crime. For example, those who received their crime news from radio, newspapers or television had [End page 110] similar levels of worry about crime. In addition, their findings indicate that crime show viewers were more likely to worry about being sexually assaulted; getting beaten up, knifed or shot; and getting killed. However, crime-drama viewing is not related to fear of car jacking, mugging, or burglary.
Researchers argue that public fear and anxiety is inextricably connected with public pressure for solutions to crime problems. A number of research studies focus on whether media depictions of crime influence public attitudes towards criminal justice policy. They find that presentations of crime news increase public pressure for more effective policing (Garofalo, 1981a) and more punitive responses to crime (Barille, 1984; Surette, 1998). Furthermore, Surette (1998:) claims that the news media feature agents of crime control as negatively ineffective and incompetent which results in support for more police, more prisons, and more money for the criminal justice system. Reith (1999) finds that for white males, crime show viewing is related to high levels of aggression towards those who break the law, and low levels against those who defend it. She also found that fear of victimization and fear of victimization based on real life experiences did not have a mediating effect on the relationship. In addition, Oliver and Armstrong (1995) find that frequent viewing and greater enjoyment of reality-based crime shows are related to holding punitive attitudes. However, frequent viewing and greater enjoyment of fictional crime shows are not related to holding punitive attitudes.
Public attitudes toward police are generally positive (Huang and Vaughn, 1996). However, there are few studies that examine the mediaís influence on public ratings of police effectiveness. Much of the literature focuses on media portrayals of police officers and findings reveal two conflicting views. Some researchers argue that the police are presented favorably in the media, while other research suggests that the police are negatively portrayed in the media.
Presentations of police are often over-dramatized and romanticized by fictional television crime dramas while the news media portray the police as heroic, professional crime fighters (Surette, 1998; Reiner, 1985). In television crime dramas, the majority of crimes are solved and criminal suspects are successfully apprehended (Dominick, 1973; Estep and MacDonald, 1984; Carlson, 1985; Kooistra et al. 1998, Zillman and Wakshlag, 1985). Similarly, news accounts tend to exaggerate the proportion of offenses that result in arrest which projects an image that police are more effective than official statistics demonstrate (Sacco and Fair, 1988; Skogan and Maxfield, 1981; Marsh, 1991; Roshier, 1973). The favorable view of policing is partly a consequence of policeís public relations strategy. Reporting of proactive police activity creates an image of the police as effective and efficient investigators of crime (Christensen, Schmidt and Henderson, 1982). Accordingly, a positive police portrayal reinforces traditional approaches to law and order that involves increased police presence, harsher penalties and increasing police power (Sacco, 1995).
In addition, a number of researchers suggest that a symbiotic relationship exists between news media personnel and the police. It is suggested that the police and the media engage in a mutually beneficial relationship. The media needs the police to provide them with quick, reliable sources of crime information, while the police have a vested interest in maintaining a positive public image (Ericson, Baranek, and Chan, 1987; Fishman, 1981; Hall et al, 1978). However, [End page 111] other researchers argue that the police are not portrayed positively in the news media. For example, Surette (1998) claims that docu-dramas and news tabloid programs represent the police as heroes that fight evil, yet print and broadcast news personify the police as ineffective and incompetent. Likewise, Graber (1980) claims that the general public evaluates police performance more favorably compared with courts and correction. Nevertheless, Graber (1980) states that the media provides little information to judge police and that the news media focus on negative criticism rather than positive or successful crime prevention efforts. In essence, most media crime is punished, but policemen are rarely the heroes (Lichter and Lichter, 1983).
Prior research suggests that public knowledge about crime and justice is largely derived from the media (Roberts and Doob, 1990; Surette, 1998). This research seeks to build on previous research by addressing three research questions:
Police effectiveness, fear of crime and punitive attitudes are important aspects of public attitudes toward crime and justice in the United States. First, police strategies reflect departmental values, which reflect community values. Negative or positive attitudes towards the police may influence police policy making and strategy. Second, citizen attitudes toward the police may influence decisions to report crime. Third, both fear of crime and punitive attitudes may influence policy making and law making by government agencies, as public support or opposition may determine policy.
The sample is derived from the 1995 National Opinion Survey on Crime and Justice (NOSCJ). The NOSCJ is a random telephone survey of adults (n= 1005) who reside in the continental United States. The survey is cross-sectional and samples are stratified to all U.S. counties in proportion to each countyís share of the telephone households in the target area. The survey employed random digit dialing (CATI) and achieved a 62% response rate. The purpose of the NOSCJ is to provide knowledge about American attitudes toward crime and justice issues, which may lead to more informed criminal justice policy and practice. The survey examines a number of issues, such as attitudes toward courts, police, neighborhood problems, juvenile gangs, drug laws, death penalty, gun control, prisons, and worries about crime. In addition to basic demographic characteristics, NOSCJ captures information about hours of television viewing, crime show viewing and source of crime news. [End page 11]
Fear/Worry of Crime
Fear of crime is measured using seven items that examine the respondentís fear/worry toward crime. Respondents are asked if they worry about sexual assault; car-jacking; getting mugged; getting beaten up, knifed or shot; getting murdered; being burglarized while at home; and being burglarized while no one is at home. Each question on worry/fear of crime has a four-category response ranging from very frequently, somewhat frequently, seldom, and never. The seven items are scaled to establish an index of fear of crime that ranges from seven (low worry) to twenty-eight (high worry). Higher scores indicate a greater amount of fear/worry about crime. Reliability analysis reveals an alpha of .86, which indicates that the scale is highly consistent.
Perception of Police Effectiveness
Police effectiveness is measured by using seven items that examine the respondentís attitudes towards police. Three questions address respondentís confidence in police ability to protect, solve, and prevent crime. Each question has a four-category response ranging from a great deal, some, little, and none at all. Three questions address respondentís assessment of police promptness, friendliness, and fairness. Each question has a five-category response ranging from very high, high, average, low, and very low. For the scaling purposes, very low and low were combined into one category. The final question examines the respondentís belief in the use of excessive force by police in their community. The category response range from serious problem, somewhat of a problem, minor problem, and not a problem at all. The seven items are scaled to establish an index of perceived police effectiveness that range from seven to twenty-eight. Higher scores indicate positive appraisals towards police effectiveness and lower scores indicate negative appraisals of political effectiveness. Reliability analysis reveals an alpha of .83, which indicates this scale is consistent.
Punitive Justice Attitudes
Punitive justice attitudes are measured by using 11 items. These questions were categorical in nature and for scaling purposes they were dummy coded. Table one presents the items that are scaled to identify those who hold punitive attitudes toward crime and justice.
Table 1: Punitive Attitudes Scale
[End page 113]
The scale ranges from 0 (punitive attitudes) to 1 (non-punitive attitudes). The scores range from 0 (highly punitive) to 11 (non-punitive) and the average score for respondents is four. The alpha level of punitive attitude scale is .72, which indicates that this scale is reliable. Nevertheless, one limitation is equating punitive attitudes with retributive attitudes. Historically, the notion of retribution meant "an eye for an eye" and emphasized "harsh" punishment. However, the concept of retribution has evolved and includes the concept of just deserts. Just deserts require that the nature of punishment be consistent with the offenderís criminal conduct. The central principle of just deserts is proportionality; the severity of the punishment should be proportional to the gravity of the offense. Punitiveness is more concerned with the prevention and reduction of crime through deterrence principles (Von Hirsch, 1998). However, it is unclear whether survey respondents understand the differences between punitiveness and retribution.
The media variables include crime-show viewing, television hours and crime news source. Crime-show viewing is measured by asking respondents if they are frequent viewers of a television crime show. Television hours are measured by asking respondents how many hours of television they watched per week. Finally, respondents were asked the primary source of crime news. The categories include television, newspaper, radio, and friends/neighbors and are dummy coded for the analysis. Specifically, the intention is to examine the print mediaís effect on fear of crime, punitive justice attitudes and perceived police effectiveness.
Socio-Demographic Measures/Control Variables
A number of control variables are employed in this research to ensure that media effects are properly measured. Demographic variables such as race, gender, age, income, residence, level of education, and marital status are employed in the analysis. Race, income, residence, level of education and marital status are dummy-coded. In addition, a scale is created to measure respondentís attitudes toward problems in their neighborhood. Respondents were asked to rate the seriousness of a number of issues in their neighborhood. The issues include: trash and litter; loose dogs; unsupervised youth; graffiti; vacant houses; noise; people drunk/high in public; and abandoned cars. The scores range from eight to thirty-two. Higher scores indicate high levels of problems in the neighborhood, whereas lower scores indicate low levels of problems in the neighborhood. Reliability analysis reveals an alpha of .81, which indicates a consistent scale.
The analytic strategy is to examine the relationship between media variables and fear of crime, perceived police effectiveness and punitive justice attitudes. The first step is to conduct univariate and bivariate analysis. The next step is to employ multivariate regression models using the ordinary least squares. Included in the models are the socio-demographic variables/control variables described above. Three models will be developed to examine the dependent variables, which will include fear of crime, punitive justice attitudes and perceived police effectiveness. The first model will examine the association between crime-show viewing, newspaper as primary source of crime news, hours of television per week and fear of crime. The [End page 114] control variables will include age, race, residence, marital status, income, gender, problems in neighborhood, and perception of police effectiveness.
The second model will examine the association between crime-show viewing, newspapers as primary source of crime news, hours of television viewing and punitive justice attitudes. We will employ the same control variables as step one, except that we will include fear of crime as an independent variable. The final step is to examine the association between crime-show viewing, newspapers as primary source of crime news, hours of television viewing and perceived police effectiveness. We will employ age, race, residence, marital status, income, gender, neighborhood problems, fear of crime and punitive attitudes as control variables.
Univariate and Bivariate Analysis
Table two presents descriptive statistics of the variables employed in this study.
Table 2: Descriptive Statistics and Correlations
P*< .05; **P< .01
The results indicate that respondentsí average approximately 15 hours of television per week, while 42% of the respondents report that they are regular viewers of crime shows, and that 20% of the respondents report that newspapers are their main source of crime news. The scaled variables were employed as both dependent and control variables. The results indicate that on a scale of seven to twenty-eight, the respondentís average score is 13.65 for fear of crime and 15 for perceived police effectiveness. On a scale of eight to thirty-two, the respondents score 11.8 [End page 115] for perceived problems in their neighborhood. On a scale from 0 to 11, respondents mean score is 4.09 for punitive attitudes toward crime and justice.
Socio-demographic characteristics of the sample indicate that 7.7% of the respondents are African-American, 7.7% are Hispanic and 81.4% are white; 53.1% are married; 52.1% are male; 15.9% are urban residents; 37; the average age is 45; 58.2% are college educated; 22.1% have incomes over $60,000; 37.2% have incomes between $30,000 and $60,000; 25.8% have incomes between $15,000 and $30,000; and 14.9% have incomes lower than $15,000.
The results of correlation analysis are also included in table 2. The results indicate that viewing crime shows is significantly related to fear of crime and perceived police effectiveness. Regular viewers of crime shows are more likely to fear or worry about crime. Similarly, regular crime drama viewers are more likely to hold negative attitudes toward police effectiveness. The bivariate analysis indicates that newspaper as primary source of crime news and hours of television viewing are not significantly related to fear of crime, punitive attitudes or perceived police effectiveness.
In addition, the results indicate that white, married, and low-income (15k to 30k) respondents are more likely to have punitive attitudes, whereas black, college educated, and respondents with low appraisals of police effectiveness are less likely to have punitive attitudes. The results also indicate that older respondents, males and respondents with low perception of neighborhood problems are more likely to have low fear of crime, whereas, younger respondents, female, Hispanic, college-educated and respondents with low appraisals of police effectiveness are more likely to fear crime. Finally, bivariate results suggest that Hispanic, African-American, urban, and younger respondents are more likely to have negative or low appraisals of police effectiveness. Conversely, respondents with punitive attitudes, with a medium income (30k to 60k), older, white, with low perceptions of neighborhood problems are more likely to have positive or high appraisals of police effectiveness. However, there may be a number of factors that mitigate or enhance the relationships. Thus, it is necessary to conduct multivariate techniques to further address these relationships
Fear of Crime
Table three presentsí three models based on OLS regression, employing fear of crime, punitive attitudes, and perceived police effectiveness regressed on media/control variables.
The findings indicate that crime-show viewing is related to fear of crime. Respondents who report that they are regular viewers of crime shows are more likely to be fearful of crime. This is true even when we control for age, gender, race, income, education, marital status, perceived police effectiveness and perceived neighborhood problems. However, hours of television and newspaper as the primary source of crime news are not significantly related to fear of crime. [End page 116]
Table 3: Multivariate Models: Fear of Crime, Punitive Attitudes and Police Effectiveness
P*< .05; **P< .01
In this model, the strongest relationship is perceived problems in the neighborhood, followed by gender, education, regular viewing of crime shows, age, income and perceived police effectiveness. Respondents who claim that there are a high number of problems in their neighborhood are more likely to fear crime. This is not surprising, as respondents may feel unsafe in an area that they believe is conducive to crime. Female respondents are also more likely to fear crime. This is consistent with prior research that shows that females are more likely to fear or worry about crime (Garofalo, 1981b; LaGrange and Ferraro, 1989; Parker, 1993; Parker and Ray, 1990; Warr, 1984 Skogan and Maxfield, 1981). College educated respondents are more likely to be fearful of crime. This result is unanticipated, as we would assume that higher education would inform subjects about the nature of crime and justice. However, college educated respondents may feel that they have more to "lose" if they are victimized. Moreover, regular viewers of crime drama are more likely to fear crime. Television portrayal of crime and justice is largely sensational, violent and fear producing. Viewers may receive a "distorted" image of the typical crime or criminal, which may produce fear or anxiety about criminal activity. Compared to respondents with average incomes (30k to 60k), lower income respondents are more likely to fear crime. This is consistent with prior research, which reveals that low-income individuals are more likely to fear crime (Will and McGrath, 1995; Skogan and Maxfield, 1981; Baumer, 1978).
Older respondents are less likely to fear crime, which is not consistent with prior research (Baldassare, 1986; Garofalo, 1981b; Skogan and Maxfield, 1981; Yin, 1980). Finally, respondents who gave poor ratings of police performance are more likely to be fearful of crime. [End page 117] These respondents may believe that police are not effectively protecting the public or their community.
Punitive Justice Attitudes
Table three presents the results of punitive attitudes regressed on the media consumption. The findings indicate none of the media consumption variables are related to punitive attitudes. The strongest indicator of punitive attitudes is race, followed by education, income, fear of crime, and marital status. African-American respondents are more likely to hold non-punitive attitudes. This may be the result of inequalities of the justice system. For example, compared to whites, African-Americans are more likely to receive harsher punishments (such as the death penalty) and African-Americans are disproportionately over- represented in prisons (Reiman, 1998). Some African-Americans may feel threatened by a punitive justice model or feel that a punitive justice model reinforces discrimination and persecution of African-Americans.
In addition, respondents with college education are more likely to hold non-punitive attitudes. Those with education may be more likely to recognize the inequalities of the justice system and determine that solutions to the "crime problem" may be better served by policies of reintegration or rehabilitation. Furthermore, compared to average income respondents, low-income respondents (15,000 to 30,000) are more likely to hold punitive attitudes towards crime and justice. This is in contrast to the lowest income respondents ($15,000 or less) who hold non-punitive attitudes. One reason for the difference may be that low income (15,000 to 30,000) respondents are more likely to bear the brunt of crime and unlike the lowest income ($15,000 or less) respondents they may feel that they have more to "lose" by victimization. As a result, low- income respondents may believe that a punitive ideology is necessary to prevent and reduce crime in the areas in which they live. Moreover, respondents with a high fear of crime are more likely to have punitive attitudes. Fear of crime may provide impetus for support of "get tough" crime policies. Finally, married respondents are more likely to have punitive attitudes. Married respondents might believe that they have more to lose if they are victimized (i.e. family and partner) and support tougher policies toward crime.
Perceived Police Effectiveness
Table three examines perceived police effectiveness regressed on media and control variables. The findings reveal that none of the media variables are related to respondentís perceptions of police effectiveness. A possible explanation is that there is little agreement on the role that police play on television crime dramas and news reports. Some research suggests that police are positively portrayed while others show that the police are negatively portrayed. However, the results indicate that age, perceived problems in the neighborhood, fear of crime, and race are significantly related to perceived police effectiveness. Older respondents are more likely have high ratings of police effectiveness, whereas younger respondents are more likely to have low ratings of police effectiveness. This is consistent with prior research that shows that compared to younger persons; the elderly have more favorable attitudes toward police (Garofalo, 1977; Hindelang, 1974; Thomas and Hyman, 1977).
Respondents who believe that there a high number of problems in their neighborhood are more likely to rate police effectiveness as being poor. Respondents may not believe that the local police are not properly fulfilling their role in the community. Similarly, respondents who have a [End page 118] high fear of crime are more likely to give poor ratings to the police. These respondents may feel that the police are not adequately protecting their communities.
Finally, African-Americans are more likely to hold low ratings of police effectiveness. This is similar to prior research which suggests that African-Americans have an antagonistic view of police (Garofalo, 1977) and that there is a "climate of distrust:" between African-Americans and law enforcement (Jacob, 1971). However, other studies indicate that residence and social class mitigate the effect of race. For example, Kusow, Wilson and Martin (1998) find little support that African-Americans are less satisfied with police effectiveness. They find that both African-American and white suburbanites are more satisfied with police performance than African-American and white urban residents. In addition, Albrecht and Green (1977) find that low-income African-Americans living in inner cities possess the least favorable attitudes toward the police. Similarly, Parker, Onyekwuluje and Murty (1995) find that African-Americans who reside in high crime areas, and who have low incomes are more likely to have held negative attitudes toward the police. Nevertheless, controlling for income and residence we find that African-Americans are significantly more likely to hold unfavorable attitudes toward police. Waddington and Braddock (1991) find that African-Americans believe that whites receive preferential police treatment and that African-Americans are subjects of discrimination. Research indicates that there is a significant association between being black and being harassed by police (Browning et al. 1994). Other factors may include an increased awareness of police corruption, racism, brutality and racial profiling. A number of significant "social" events occurred during the 1990's. For example, the beating of Rodney King and the racist remarks of Mark Fuhrman elevated racism and police brutality into national issues. Finally, we have seen racial profiling or "driving while black" emerge as an important social issue.
This study reveals that regular viewers of crime shows are more likely to fear crime. Although statistically significant, the strength of this finding is minimal. In addition, there are a few limitations with regard to the measures of media consumption. First, the type of crime show that the respondent is viewing is unknown. There are numerous types of crime shows that may focus on different aspects of the criminal justice system. For example, crime shows may focus on police, courts, private investigators, defense lawyers and sometimes even the criminals. In addition, some shows are more realistic, while others routinely portray violence, and consistently misinform viewers about the nature of the criminal justice system and criminality. It would be prudent to know which dramas the respondents are viewing. Second, employing television hours watched is problematic, since there is no way of determining what type of programs the respondent is viewing. There are a number of different programs that may or may not address criminal justice issues and address them in substantially different ways. Finally, examining newspapers as the primary source of crime news suggests that only newspapers influence respondents. It would be naive to suggest that respondents are not affected by a number of sources; for example, respondents who receive their primary crime news from newspapers may also be affected by presentations of crime from other sources such as films, television and/or personal experiences. [End page 119]
Fear of Crime
Despite these limitations, there are some interesting results regarding fear of crime and perceived police effectiveness. Even when controlling for a number of factors, viewing crime shows is weakly related to fear of crime. Fear of crime may be "natural" reactions to the violence, brutality, and "injustice" that are broadcast to living rooms on a daily basis. Crimes on television shows and films reveal several trends. There is an overemphasis on crimes of violence and offenders are often portrayed in stereotypical ways. For example, murder and robbery dominate while property crimes are rarely presented (Surette, 1998). Offenders are often viewed as psychopaths that prey on weak and vulnerable victims. In other cases offenders are portrayed as businessmen or professionals that are shrewd, ruthless, and violent. Television crime is exciting and a rewarding endeavor, whereas victims are passive, helpless and vulnerable (Surette, 1998).
Many viewers may not understand the justice process and are unlikely to understand motivations and causes of criminal behavior. The criminal justice system is portrayed as largely ineffective, with the exception of a few "heroes" that provide justice or in some cases vengeance towards offenders (Surette, 1998). Crime shows rarely focus on mitigating issues of criminal behavior and are unlikely to portray offenders in a sympathetic or even realistic fashion. On television, crime is freely chosen and based on individual problems of the offender. Analysis of crime dramas reveal that greed, revenge and mental illness are the basic motivations for crime and offenders are often portrayed as "different" from the general population (Lichter and Lichter, 1983; Maguire, 1988) Thus, viewers may believe that all offenders are "monsters" to be feared. Consequently, heavy viewers may perceive crime as threatening, offenders as violent, brutal or ruthless and victims as helpless. These inaccurate presentations, as well as the portrayal of crime as inevitable/non-preventable may lead to an increase in the fear of crime. Nevertheless, the relationship between fear of crime and crime show viewing is statistically weak. As a result, it is important for future research to examine the relationship by employing triangulated strategies such as content analysis, experimental and survey research designs.
The results indicate that perception of police effectiveness is not related to media consumption. However, African-Americans and respondents who report a high number of problems in their neighborhood are more likely to give negative evaluations of police effectiveness. Therefore, direct experience may influence the respondentsí attitudes toward crime problems and police response in the neighborhood. Future research should examine how the media influences these attitudes. The media may produce "feelings" that local neighborhoods are "problem filled" or dangerous. For instance, local news broadcasts may focus on highly sensational, violent and disturbing crime that occurs in the neighborhood. It may be possible that media presentation will affect attitudes toward the neighborhood.
In this sample, African-Americans are more likely to give poor ratings of police effectiveness. However, it is unclear as to why or how African-Americans gain these views. It is generally assumed that these views are the result of discrimination. Direct experience aside, the mass media may play a role in African-American attitudes toward police effectiveness. Future studies should examine how the media portrayal of the criminal justice system affects African-American [End page 120] attitudes toward police. The media may have a strong effect on African-American criminal justice attitudes.
In conclusion, it is speculated that the majority of the publicís knowledge about crime and justice is formed through media consumption. As a result, it is imperative that we understand how the media influences public attitudes. Although there are limitations within the data set and the findings are weak, regular viewing of crime shows is related to fear of crime. However, crime show viewing is not related to punitive attitudes or perceived police effectiveness, while hours of television viewing and source of crime news are not related to fear of crime, punitive attitudes or perceived police effectiveness. Nevertheless, more research is required to determine the relationship between media consumption and attitudes toward crime and justice.
 These questions are reverse coded.
 With the exception of police force, the six items are reverse coded.
 Responses are dummy coded; Regular Viewer 0, Not a Regular Viewer 1.
 Originally, education was measured as last grade attended and divided into eight categories which included: 0-4; 5-8; some high school; high school graduate; some college; college graduate; and Graduate work. This was recoded into college (0) and no college (1).
 The variables are reverse coded.
 The exception is perceived problems in neighborhood, which is only employed as a control variable.
 The mean was inserted for cases with missing data. VIF and tolerance
statistics indicated no problems with multi-collinearity. Case-wise
diagnostics revealed several outliers which were excluded from the analysis.