<%BANNER%>

A Bloody Weekend in Manaus

Permanent Link: http://ufdc.ufl.edu/UFE0044202/00001

Material Information

Title: A Bloody Weekend in Manaus a Case Study of Local TV in Amazonas, Brazil
Physical Description: 1 online resource (121 p.)
Language: english
Creator: Mourao, Rachel Reis
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: brazil -- media -- television -- victimization
Latin American Studies -- Dissertations, Academic -- UF
Genre: Latin American Studies thesis, M.A.
Electronic Thesis or Dissertation
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )

Notes

Abstract: The idea that the frequency and types of crime and violence depicted in the media bear little relationship to the actual risks of victimization will come as no surprise to most people, yet few studies in Latin America have systematically documented the difference between the two. This study addresses the issue by comparing the results of a content analysis of crime-related television programs in Amazonas, Brazil, to estimates of actual crime victimization rates derived from an analysis of Brazil’s 2009 National Household Survey. Additional information comes from in-depth interviews with producers, journalists, and viewers. Together analyses of these data allow me to answer four questions:1) What is the actual prevalence of crime in Amazonas, and how does it vary by social status? 2) What is the profile of the media-constructed image of the prevalence and character of crime and violence? 3)In what ways do the actual estimates differ from the image the media presents to the public? 4)What decision-making processes do newsrooms employ when they cover crime in the state? Answers to these questions provide insight into the media’s role in constructing social realities that generate stereotypes, promote moral panics, and influence people’s political preferences.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (M.A.)--University of Florida, 2012.
Local: Adviser: Wood, Charles H.
Statement of Responsibility: by Rachel Reis Mourao.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044202:00001

Permanent Link: http://ufdc.ufl.edu/UFE0044202/00001

Material Information

Title: A Bloody Weekend in Manaus a Case Study of Local TV in Amazonas, Brazil
Physical Description: 1 online resource (121 p.)
Language: english
Creator: Mourao, Rachel Reis
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: brazil -- media -- television -- victimization
Latin American Studies -- Dissertations, Academic -- UF
Genre: Latin American Studies thesis, M.A.
Electronic Thesis or Dissertation
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )

Notes

Abstract: The idea that the frequency and types of crime and violence depicted in the media bear little relationship to the actual risks of victimization will come as no surprise to most people, yet few studies in Latin America have systematically documented the difference between the two. This study addresses the issue by comparing the results of a content analysis of crime-related television programs in Amazonas, Brazil, to estimates of actual crime victimization rates derived from an analysis of Brazil’s 2009 National Household Survey. Additional information comes from in-depth interviews with producers, journalists, and viewers. Together analyses of these data allow me to answer four questions:1) What is the actual prevalence of crime in Amazonas, and how does it vary by social status? 2) What is the profile of the media-constructed image of the prevalence and character of crime and violence? 3)In what ways do the actual estimates differ from the image the media presents to the public? 4)What decision-making processes do newsrooms employ when they cover crime in the state? Answers to these questions provide insight into the media’s role in constructing social realities that generate stereotypes, promote moral panics, and influence people’s political preferences.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (M.A.)--University of Florida, 2012.
Local: Adviser: Wood, Charles H.
Statement of Responsibility: by Rachel Reis Mourao.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0044202:00001


This item has the following downloads:


Full Text

PAGE 1

1 A BLOODY WEEKEND IN MANAUS: A CASE STUDY OF LOCAL TV IN AMAZONAS, BRAZIL By RACHEL REIS MOURAO A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGR EE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Rachel Reis Mourao

PAGE 3

3 To my sister Renata, who hates crime shows

PAGE 4

4 ACKNOWLEDGMENTS It is a pleasure to thank those who made this thesis possible. First and foremost, I thank the crew fro m Al Amazonas and Comunidade Alerta for letting me be part of their group. This thesis would not have been possible without the help from Joo Artur Vi eira and Miguel Mouro. I would also like to thank my coders Cayo Costa and Camilla Gonalves, who coded hours and hours of crime shows and got paid in ramen noodles. Special thank s goes to Dr. Cory Armstrong, Dr. Charles Wood, and Dr. Barbara Zsembik for their enthusiasm and immense knowledge. Finally, I would like to thank my Malas friends, especially Adr ian Zeh, for sharing this journey with me. It has been an honor. Obrigada!

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 10 .. 12 Comunidade Alerta ................................ ................................ ................................ 14 Al Amazonas ................................ ................................ ................................ ......... 17 Research Questions ................................ ................................ ............................... 21 What Follows ................................ ................................ ................................ .......... 21 2 ME DIA AND CRIME ................................ ................................ ............................... 24 ................................ ................................ ............................... 25 ................................ ................................ ..................... 31 3 CRIME ON THE STREETS ................................ ................................ .................... 38 Lethal Victimizations ................................ ................................ ............................... 39 Non lethal Victimizations ................................ ................................ ......................... 44 General Risk ................................ ................................ ................................ ..... 45 Theft ................................ ................................ ................................ ................. 48 Robbery ................................ ................................ ................................ ............ 50 Assa ult ................................ ................................ ................................ .............. 51 4 CRIME ON TV ................................ ................................ ................................ ........ 59 Sampling ................................ ................................ ................................ ................. 60 Coder Training and Reliabili ty ................................ ................................ ................. 61 Unit of Analysis ................................ ................................ ................................ ....... 62 Coding Main Crime ................................ ................................ ................................ 62 Coding Victims ................................ ................................ ................................ ........ 63 Coding Demographic Characteristics ................................ ................................ ...... 63 Results ................................ ................................ ................................ .................... 64 5 CONCLUSION ................................ ................................ ................................ ........ 81

PAGE 6

6 APPENDIX A CODING GUIDE: PORTUGUESE ................................ ................................ .......... 91 B CODING GUIDE: ENGLISH ................................ ................................ ................. 103 C PICTURES ................................ ................................ ................................ ............ 114 LIST OF REFERENCES ................................ ................................ ............................. 117 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 121

PAGE 7

7 LIST OF TABLES Table page 3 1 Victimization rates per I,000 inhabitants Amazonas, Brazil 2008/09 ............. 53 3 2 Victimization rates per I,000 inhabitants Characteristics of the victimizations, Amazonas, Brazil 2008/09 ................................ ....................... 53 3 3 Victimization rates per 1,000 inhabitants Characteristics of the victims, Amazonas, Brazil 2008/09* ................................ ................................ ............. 55 3 4 Racial composition Amazonas, Brazil 2008/09 ................................ .............. 56 4 1 Number and percentage of stories that depicted offense types by program June 2011 ................................ ................................ ................................ ........... 77 4 2 Use of violent images per story by show June, 2011 ................................ ....... 77 4 3 Number and percentage of TV stories that depicted vi ctims characteristics by program June, 2011 ................................ ................................ ........................ 78 4 4 Number and percentage of stories that depicted non lethal victims char acteristics* June,2011 ................................ ................................ ................ 79 4 5 Portrayal of the victim by race ................................ ................................ ............ 80 4 6 Portrayal of victimization by gender ................................ ................................ .... 80 A 1 Images Portrayed Portuguese ................................ ................................ ....... 101 A 2 Im ages portrayed English ................................ ................................ .............. 112

PAGE 8

8 LIS T OF FIGURES Figure page 2 1 Media, process, and the perception of crime. ................................ ..................... 37 3 1 Racial composition Amazonas, Brazil, 2008/09. ................................ .............. 57 3 2 Non lethal occurrences by race (rate per 1,000 inhabitants) Amazonas, Brazil, 2008/09 ................................ ................................ ................................ .... 57 3 3 Non lethal occurren ces by age (rate per 1,000 inhabitants) Amazonas, Brazil, 2008/09 ................................ ................................ ................................ .... 58 3 4 Non lethal occurrences by income (rate per 1,000 inhabitants) Amazonas, Brazil, 2008/09 ................................ ................................ ................................ .... 58 B 1 Al Amazonas crew interviewing criminal in police station in Manaus 06/14/2011. ................................ ................................ ................................ ....... 114 B 2 Reporter from Al Amazonas meeting with police. ................................ ........... 115 B 3 Al Amazonas equipment mixed with crime evidence. ................................ ..... 115 B 4 Host of Comunidade Alerta discussing a homicide with Raimundinho, the uppet ................................ ................................ ................................ ... 116

PAGE 9

9 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts A BLOODY WEEKEND IN MANAUS: A CASE ST UDY OF LOCAL TV IN AMAZONAS, BRAZIL By Rachel Reis Mourao May 2012 Chair: Charles Wood Major: Latin American Studies The idea t hat the frequency and types of crime and violence depicted in the media bear little relationship to the actual risks of victim ization will come as no surprise to most people yet few studies in Latin America have systematically documented the difference between the two This study addresses the issue by comparing the results of a content analysis of crime related television progr ams in Amazonas, Brazil, to estimates of actual Survey. Additional information comes from in depth interviews with producers, journalists, and viewers. Together analyse s of these data allow me to answer four questions: What is the actu al prevalence of crime in Amazonas and how does it vary by social status? What is the profile of the media constructed image of the prevalence and c haracter of crime and viole nce? I n what ways do the actual estimates differ from the image the media presents to the public ? What decision making processes do newsrooms employ when they cover crime in the state? Answers to these questions provide insight into the social realities that generate stereotypes, promote moral panics, and influence political preferences.

PAGE 10

10 CHAPTER 1 INTRODUCTION young man hours of inquiry, frustration took over. No one would comment on the case, later found to be a homicide perpetrated by a local drug dealer. With no information to go on, the report spent under the blistering Amazonian sun, the TV crew of Al Amazonas finally gave up. Just as the crew was about to leave, the reporter decided to question the father of t he victim one last time. Persistence was finally rewarded when the father revealed that it was his anti drug evangelical campaign that had seriously angered the drug dealers who ran the area Not far away, in the studios of Band Amazonas, Ronaldo Tabosa an d his son prepare to go live on local television with their show Comunidade Alerta. The general atmosphere in the studio is cheerful. Producer and cameramen chat about soccer the day before, Vasco da Gama had won the Copa do Brasil after ten years withou t a title. broadcasts his image inside the houses of thousands of Amazon citizens during their l unchtime. What follows is a 30 minute show portraying the arrest of a gang by the Military Police and the death of a teenager by his cocaine addicted brother as part of a satanic ritual. By 12:30, TV A Crtica starts broadcasting its most famous show, Al Amazonas, a newscast dedicated to stories about crime and denunciations in low income communities. The news of the day includes three robberies, the death of a cab

PAGE 11

11 driver due to involvement with drug trafficking, and the teenager satanic homicide. When I q uestioned Tabosa as to why he decided to create a crime show, his answer truth? How represe ntative is the portrayal of crime in the State of Amazonas of the actual risk of being victimized? Far from the slums of Rio de Janeiro and the international drug trafficking conflicts that so often make it to the international headlines, Amazonas viewers constantly watch local shows which claim that the state is dominated by drug trafficking, that homicide These stories are often accompanied by images of drugs, weapons, corpses, and Comunidade Alerta only another example of the adage in journalism it With these questions in mind, I spent six weeks in Manaus interviewing creators, producers, directors, reporters, viewers and key informants on the issue. I also accompanied the reporting teams and was present in the studios for the live broadc asting of the shows, as well as during their meetings before and after. This provided me a unique opportunity to see media practitioners interacting with criminals in jail, the family of victims at funerals, police commissioners in police stations, and wit nesses at crime scenes, among others. For several weeks, I arrived at the TV station

PAGE 12

12 routine of calling each police station in the city to the shootings and live hits. Us ing recorded episodes, I analyzed the content of two shows, both of which are broadcasted from Manaus to a majority of the Amazonas state. Identified in 2009 as media (New York Times, CBS, Associated Press), Manaus is the seventh most populous city in Brazil and had a rate of 32.5 homicides per 100,000 inhabitants in 2009 (Waiselfisz, 2010). Geographically distant from media centers, Amazonas rarely makes it to national bro adcasts. As a consequence, local TV shows and newscasts are the only gatekeepers of information about what happens in the state. the Development of P op ular Crime S hows in Brazil Comunidade Alerta and Al Amazonas form part of t he growth in crime shows which began in Brazil in the early 1990s. The first successful show of this type to achieve nationwide appeal was called Aqui, Agora. Broadcasted by the Sistema Brasileiro de Televiso (SBT) and aimed at low income viewers, the sho w introduced techniques like direct camera (where the host directly faces the viewers) and long takes using a dolly or a Steadicam to reproduce the action in real time. Aqui, Agora achieved remarkably high ratings for a television show outside of the Globo achieving a 31 percent share in Sao Paulo by 1992 (Hamilton, 2009). In response, other networks quickly copied the format of the show, blending criminal reports with community discontents complaints about lack of public services in general, like water, electricity, bad roads, etc to produce their own variations of Aqui, Agora. Currently, Band Jos Luiz Datena, the personificati Hosts lik e Datena

PAGE 13

13 mediate the relationship between society and government, denounce crime and call on authorities to act. Along the same lines, Rede Record broadcasts the show Reporter Record. These shows air nationally from 3:30 to 6:30 p.m. and compete with one o f running soap operas, the teenage drama Malhao. The crime show phenomenon rapidly spilled over into local television in Brazil. informative local newsc asts, its competitors invested in a more opinion based representation of crime and airing community complaints using popular language. with a language appropriate for the audience they address. They are, therefore, directed The year 2004 marked the golden age of crime shows in Amazonas, wihen local networks broadcasted more than six daily shows dedicated to crime and community. The shows were a udience leaders, as they responded to the perception that the government was not addressing the problems that afflict the poor (Gusmo, 2004). The popularity of crime shows was also evident in the 2004 elections. Sabino Castelo Branco, host of Bronca na TV and Wallace Souza, host of Canal Livre, were the most Carlos Souza, also host of Canal Livre, was the most voted Federal Deputy in the history of Amazonas, with 147, 2 12 votes, or 10 percent of the total electorate. Sabino and the Souzas owned crime shows and, with the support of traditional local politicians, presented themselves as vigilantes. They invaded slums and gave arrest orders, becoming stars in the process, p laying the role of

PAGE 14

14 (Correa, 2004). However, in 2008, the biggest scandal in the history of the Amazonas media put an end to the dominance of crime shows in local TV and politics. Wallace Souza, host of Canal Livre and sta te legislator, was on the cover of the main national and international newspapers, accused of plotting murders to boost TV ratings. Wallace and his son faced charges of eliminating political rivals and commanding a network of organized crime in the state. They were accused of homicide, drug trafficking, illegal gun possession, and attempting to destabilize public security in Amazonas (Associated Press, 2009). In 2009, Wallace lost his mandate. He was arrested in October of 2009, but soon transferred to a h ospital in So Paulo due to a chronic liver condition. After a prolonged stay in the hospital, Wallace died of a heart attack in July of 2010. Both Canal Livre and Bronca na TV were cancelled and, although Sabino and Carlos Souza still hold positions in th e municipal legislature, the only crime shows that survived this period were Al Amazonas and Comunidade Alerta. Comunidade Alerta Comunidade Alerta is an independent production broadcasted by Band Amazonas and employs a mixed format of talk show and new s stories focusing on crime. The show airs daily from 11:30 a.m. to 12:00 p.m. and is hosted by Ronaldo Tabosa and his son, Jander Tabosa. Ronaldo Tabosa ran and lost the elections for state legislator in 2010. Jander won a city council position and was ex pelled soon after due to charges of electoral fraud. In addition to the hosts, the show also has a puppet, called Raimundinho, who provides funny comments on the events of the day, trying to lighten up the otherwise somber atmosphere in the studios. Comu nidade Alerta is an independent production. Its reporters and producers are not part of Band Amazonas staff. Instead, they work for Ronaldo Tabosa, who buys a

PAGE 15

15 minute daily show in Band Amazonas costs approximately R$25 ,000 (equivalent to U$14,000) per month, plus the production costs. All the advertising revenues go to the owner of the show. This type of independent production is not uncommon in Amazonas. In fact, Band Amazonas, former TV Rio Negro, currently broadcast s other two shows created and host by politicians (Camera 13, hosted by congresswoman Conceio Sampaio and her sister, city council Socorro Sampaio; Exija seus Direitos, hosted by state deputy Marcos Rotta). According to Costa (2008): In Amazonas, the ne twork Tv Rio Negro stood out because of its political alliances and because of the number of hosts that have been part of the City Hall [ Cmara Municipal de Manaus ], State Assembly [ Assemblia Legislativa do Amazonas ], and Congress [ Cmara dos Deputados ]. The businessman Francisco Garcia, founder of Tv Rio Negro and currently the president of Partido Progressista (PP), is also the main articulator of the political alliances in the last twenty two years. In August 2008, Tv Rio Negro was bought by Grupo Bande irantes de Comunicao, becoming Band Amazonas. However, the political influence of PP and the Garcias is still present, as well as the host politicians that own TV shows similar to Comunidade Alerta. Comunidade Alerta was created in 2004, and started as a community program, where people could call attention to the problems in their community on television and demand that authorities take action. Over the years, Comunidade Alerta slowly became a crime show. According to Ronaldo Tabosa (personal communicatio n, June 7, 2011), the show started to focus on crime due to the increase in violence in the city of Manaus a public service:

PAGE 16

16 We try to show the criminal. We show their faces as much as we can, so people who have been victimized by them before can report crimes in the police. Criminals should be in jail, and the public security services are not enough to provide safety in Manaus. (R. Tabosa personal communication, June 7, 2011) When asked about the charges that he used the show for personal political benefit, he replies: I was a candidate because the population kept asking me to do that. However, if that was the goal of the program, after I lost the elections, I would h ave given up the show. This is not the case. O ur show is not politics oriented. With respect to the images used, Comunidade Alerta initially aired during a late night time slot, when they could broadcast uncensored images of crime scenes. However, since 2 008, the show moved to lunchtime and had to adapt to the broadcast rating policies, and extremely graphic images (corpses, mutilations) are now blurred. The show still shows images of criminals, weapons, drugs and blood. The daily routine of the reporters of Comunidade Alerta starts early, with a route, called ronda or patrol, on the police stations of the city. The reporters have their sources, but most of the stories come from anonymous phone calls. The role of anonymous phone calls is essential to the sh scoops from the streets, but the audience can also go on air to protest and express indignation. This feature makes the show interactive and personal. According to Oliveira (2008), this type of interactivity reinforces the notion of public service already attributed to local crime shows. However, nothing is more characteristic of a local crime show in Brazil than the strong and peculiar role of the host. According to Oliveira (2008): The host [of crime shows] is the one who plays the role of mediating the relationship between the program, viewers, interviewees, and other

PAGE 17

17 characters involved in the elaboration of the show. Normally, it is the host with the audience. attention to the many risks people face. Tabosa is included in t his category of host: a mediator between the people and the authorities, a vigilante who is there to expose risks and pressure authorities for a solution. Al Amazonas Broadcasted by TV A Critica, Alo Amazonas employs a newscast format that airs crime stor ies and community grievances The show has a tradition of providing more professional coverage independent of any political group. In fact, the crew of Alo Amazonas is under the direction of the journalism sector of Rede Calderaro de Comunicao, a conglom erate that owns three newspapers, three TV channels, two radio stations, and one web portal. voi ce overs, live shots and covering current affairs. It uses the traditional language of focus on crime stories and investigative reports, but other than the major theme viol ence they resemble a regular newscast. The Al Amazonas crew begins its day when the producer calls police stations to episode and updates the list of missing people in the city of Manaus. Al Amazonas has two teams that consist of one reporter and one cameraman who work from 7 a.m. to 11

PAGE 18

18 a.m. gathering the stories of the day in the city and its surroundings. By 11:30 a.m. all stories should be in the editing room for the 12: 30 broadcast. Both reporters of Al Amazonas cover crime, but it is Fabiola Gadelha, their female reporter, who is known as one of the most respected crime journalists in the city. Gadelha has developed a personal style while covering her stories that inc ludes her signature outfits, hairstyle, and make up she does not have to follow the traditional ly discrete journalism standards and the aggressive way she interviews criminals. Fabiola Gadelha has been working as an investigative reporter for more than ten years and has been part of the crew of Al Amazonas since 2006. When describing her work, Gadelha explains: I have no limits. I do everything I think needs to be done, and when the tape lways try to show their [criminal] faces and convince the commissioners to allow me to get the best stories. If I think a story is worth it, I will do anything I can and cross any limit. Our weapon is the camera. (personal communication, June 14, 2011) Fab reporters achieve. When I accompanied her on h er daily routine, I was continuously overwhelmed by the feedback from the population, prompted by her ability to get people to talk, and by her unique way of finding a dramatic plotline from jails to crime scenes and funerals, no matter how hard the story seemed to be. When asked about the consequences of the political scandals that involved crime shows in 2004, she replied: I think inv estigative journalism in the state has suffered a lot, but not our show in particular. Since we have no politician working with us, the show actually gained audience and respect in Manaus. I believe people are calling us more, they believe in our show and in our reporters. (personal communication, June 14, 2011).

PAGE 19

19 It is important to note that although none of the producers, reporters or hosts are politicians, it is not accurate to say that the group operates without any political affiliation. In Latin Ameri ca, the government subsidizes most media outlets and, even when the companies are privately owned, they depend largely on government advertis ing (Salwen & Garrison, 1991). Picard (in Salwen & Garrison, 199 1) listed several other ways in which governments c an economically intervene in privately owned Latin America media. The list includes government investment, tax benefits, and .27). Currently, there are no official ratings of audience numbers in Amazonas. Both Band Amazonas and TV A Crtica have their internal measures, including feedback from the public through phone calls or surveys. While Band claims that Comunidade Alerta re aches 300,000 households daily, Tv A Crtica simply says that Alo Amazonas is the Institute of Public Opinion and Statistics (IBOPE), the equivalent of Nielsen media research is e xpected to provide the first official ratings and share numbers in Manaus by the end of 2011. The relationship between media and crime has been the focus of social science research in Europe and the United States since the 1970s, but very little research has studied this relationship or how the media portray crime in Brazil. The purpose of this study is twofold, based on quantitative and qualitative methods. First, using the 2009 National Household, I generated estimates of the prevalence of crime victimi zation. I then carried out a content analysis of selected television programs in order to document

PAGE 20

20 the way that the media construct public images of the prevalence and character of crime and violence. This study aims to fill the gap in the literature of c rime and media in Brazil, focusing on local television news shows. By verifying the difference between news coverage using content analysis and official victimization statistics, it was possible to obtain an estimate of the magnitude of the gap between med ia and reality, which cultivation research has shown not only to simply generate exaggerated and/or stereotyped fear on audiences, but also to resonate shared social values. Content analysis can examine content after it is produced, in a content centered model of communication. However, there are antecedent conditions that influence the content in the communication process. Currently, most content analyses are descriptive and not linked to the message production or its effects (Riffe, Lacy & Fico,2005). A ccording to Shoemaker and Reese (in Riffe, Lacy & Fico,2005), the result is that, until researchers are able t o link content to the forces that cre ated it and its effects on fe, Lacy & Fico, 2005, p.12). In order to complement the findings from the content analysis, I spent six weeks in Manaus as a participant observer of the shows production and reporting teams. h a recent ethnographic qualitative interviews with producers, key informants and viewers, I set out to understand the process of news making and news consumption of cri me shows, as evident by answers to four research questions.

PAGE 21

21 Research Q uestions Of all the issues raised by the overwhelming success of crime shows in the region, I was most interested in the difference between the characteristics of victimizations portraye d on television and the actual incidence of crime, as captured by o fficial What is the actu al prevalence of crime in Amazonas and how the m edia constructed image of the prevalence and c n what ways do the actual estimates differ from the image the media presents to What decision making processes do newsrooms employ when they cover crime in the state? In addition to those questions, based on the concept of host vigilante or xerif es eletrnicos, I will test the hypothesis that Comunidade Alerta (host politician) is more violent tha n Al Amazonas (non political) (Hypothesis 1). What F ollows This study is divided into three chapters: Media and Crime Crime on the Streets, and Crime on TV. The C hapter 2, bleeds it leads literature review on t he relationship between media and crime which has been the focus of social science research in Europe and the United States since the 1970s The on for Brazilian society. provide s estimates of victimization and risk exposure in Amazonas using the Nat ional Household Survey (PNAD). In 2009, PNAD contained a supplement with twenty four questions about victimi za tion. With this

PAGE 22

22 dataset, it was possible to measure the occurrence of nonfatal crimes reported and non reported in the state of Amazonas. Because PNAD is based on survey with victims, it is impossible to measure for ho micide. For fatal crimes, I used the d ata from Mapa da Violncia 2011 Os jovens do Brasil (Weiselfisz, 2011). Using cross tabulations, I obtained the characteristics of the occurrences and the victims of theft, robbery and assault in the state. These characteristics include: what was robbed /stolen, place of crime, relationship with assailant, a ge, sex, race, occupation, and income of the victims. ustice) Chapter 4 through a content analysis of Comunidade Alerta and Al Amazonas during the period of June 06 17, 2011. The unit of analysis is each story (approximately 5 min each). I used a data collection protocol developed by Cavender and Bond Maupin (1993) in their analysis of crime shows in the United States. The protocol focuses on three aspects of programming: demographics (general information on criminals and victims); chara view (relative safety of people, places and attitudes the audience should have) (Cave nder & Bond Maupin, 1993). Chapter 4 will be complemented with information gathered in interview s and during my participant observation in Manaus for six weeks. Chapter 5 the analyses of actual occurrence of crime and its portrayal in the media. The gap between crime as a media con struction and as a reality in Amazonas is the focus of this

PAGE 23

23 section. Limitations of this study and recommendations for future research are also included. constructing social re alities that significantly depart from the actual profile of crime and its victims. Because they generate false stereotype, promote moral panics, and viewers have profo und social consequences.

PAGE 24

24 CHAPTER 2 MEDIA AND CRIME The relationship between crime and the media, especially in the United States, Japan and Europe, has long been a topic of concern in the fields of criminology, psychology, social sciences and media stud ies. Since early years of American media development crime has always been regarded as newsworthy. According to LaMay and Dennis (1995) rime is the most common and least studied staple of news. Its prominence dates at least back to the 1830s, when the urban penny press employed violence, sex and scandal to build dizzying new levels of circulation a n d beg in the (p. 1) Several theoretical approaches have been used to study crime news: For Durkheimian functionalists, for example crime reporting can help to reinforce moral boundaries and promote social cohesion among the law abiding by publicly labeling and denouncing particular forms of criminal transgression. For Marxists, it contributes to strengthening the legitimacy of political consent around the interests of a powerful minority. For Feminists, crime news can contribute to the wider subjugation of women in a patriarchal society by reinforcing the predat ory dominance of males an d the victim status of females What all these approaches share in common, however, is their acknowledgement that news media representations of crime and criminal justice are crucial, yet highly selective and unrepresentative, source of information about crime, control and social order. (Greer, 2010, p.200) Chapter 2 reviews the main theoretical frameworks used to assess the relationship between media and crime, as well as the development of crime shows in Brazil. The first sec provides a literature review on crime coverage and its effects, as well as the importance of this study to the existing literature.

PAGE 25

25 suprem acy in Brazil. Finally, I will address the main criticisms involving the current coverage of violence in the region. The strong appeal of crime and violence to audiences suggest that media representations do not reflect real statis tics: crime on television is more violent, random and dangerous than in reality (Muraskin & Domash, 2007). The characteristics of violence on Western media are well known Jewkes (2004) describes the twelve criteria that shape crime news. They are: 1. Thresho ld: I s the story significant enough? Does it have enough drama? 2. Predictability: C an the outcome be predicted? Is it also newsworthy? 3. Simplification: C an it be simplified? Are the audiences going to be confused? 4. Individualism: S tories should have a human f ace 5. Risk: T he idea that we could all be victims of the crime, proximity 6. Sex: T he most salient news values, when associated with crime. Sexually motivated murderers are particularly attractive 7. Celebrity: I f a celebrity is involved, the crime will make it to the headlines 8. Proximity: S pecial and cultural proximity make the crime more newsworthy than if it had happened in a poorer community or a foreign country. Also related to risk. 9. Violence: T he violent crimes, as opposed to the mugging crimes, have a big ger dramatic appeal 10. Spectacle and graphic images: W ith television, the crimes covered need shocking images to entertain and give realism 11. Children: A ppeal to emotions 12. appl ied in the United States as the capacity of politicians to take advantag e of the occurrence of crimes (p. 40)

PAGE 26

26 Several studies have shown that news media do not produce content that is consistent with the actual occurrence of crime. Violent crime stories especially outrageous, violent or sexual ones seem to boost ratings. Accord ing to Surette (1998): C rime is seen to be the single most popular story element in the history of U.S. commercial television, with crime related shows regularly accounting for o ne fourth to on e third of all prime time shows (in Muraskin & Domash, 2007, p.7). product of a complex process which begi ns with a systematic sorting and selecting of events and topics according to a p.53 ). In fact, Katz (1987) defends the idea that the most newsworthy articles are not necessarily the most unexpect ed crimes. According to the author: Especially newsworthy articles do not appear to be especially unexpected, e ither to victims or to readers. In short, if crime news inevitably carries a sense of the unexpected, just what that sense is, is not obvious Perhaps, following Durkheim, we should turn the explanation around and consider whether crime is newsworthy for its symbolic value in articulating the normatively expected (Ketz, 1987, p. 230) In other words, crime stories do not cover all forms of crime and tend to exaggerate unusual crimes (Jewkes, 2004), yet not straightforward. In fact, the process of defining what constitutes the news can reveal some characteristics and shared values of a specific society. Acco rding to Katz (1987) yesterday, but because readers seek opportunities to shape up moral attitudes they will

PAGE 27

27 Ba rak (1994) proposes a model of medi a, process and t he perception of crime (F igure 2 1 ), defined as PERCEPTION OF CRIME = MEDIA x (CULTURE + POLITICAL ECONOMY) OVER TIME. According to the author: The model suggests that how we define the cultural production of crime as how we regard victims, offenders, and agents of crime control, emerges out of the social interactions between ordinary people, journalists, and sources of information within the structural and political economic contexts of active processes of news constr uction and crime management (Barak, 1994, p.6). Starting in the late 70s, scholars (Gerbner et al., 1994) tried to explain the media : Those who spend more time watching television are more likely to perceive the real world in ways that reflect the most common and recurrent messages of the television world in comparison to people that watch less television but are otherwise comparable i n terms of important demographic characteristics (Shanahan & Morgan, 1999, p. 4) In the field of criminology, the study of fear of crime has been debated since the ne tendency has been to see this fear as having no ground ings on reality but being merely the product of sensational and selective news c (p. 109). Studied that conclude that media portrayal of violence affects viewers have been subject to question mainly for the unwarranted ass umption that the audience is passive and homogeneous. Audiences carry their own background, which includes previous experiences with crime, victimization and other types of insecurities. According to Wood and Ribeiro (2011) Dammert and Malone noted that fe ar of crime [in Chile] was highly correlated with insecurities associated with employment, education opportunities for children, quality of life, economic vulnerability, and human

PAGE 28

28 scap cause, and chart a response. (no page) In the United States, research conducted by Newcomb (1978) challenged the cultivation theory in regards to fear of crime (Shanahan & Morgan, 1999). history and culture, and that all viewers do not interpret an act of violence in the same way. He defended the value of an in depth analysis of individual programs against the Cultural Indica tors [defended by Cultivation theorists] focus on aggregate patterns. Especially, he emphasized the role and extent of differences differences in context, differences in programs, differences in viewers. These differences mean that no program, much less all programs, can have a single invariant meaning that is unproblematically Shanahan & Morgan, 1999, p.60). However, Shanahan, Morgan and Signorielli (2009) do not see these differences in audience inter actions with messages as a denial of the Cultivation Theory. According to them: From the reception perspective, it seems logical to argue that other circumstances do intervene and can neutralize the cultivation process, that viewers do watch selectively, t hat programs make a difference, and that how viewers construct meaning from texts is more important than what they watch. We do not dispute these contentions; the polysemy of mediated texts is well established. From the cultivation perspective, though, to say diversity and complexity does not negate that there can be important commonalities and consistencies across large bodies of media output. To explore those commonalities, as cultivation does, is not to deny that there are differences; similarly, the examination of differen ces need not (and arguably, can not) deny the possibility of shared meanings in a culture (Morgan, Shanahan & Signorielly, 2009, p.37). Shanahan and Morgan (1999) contend that this cultivation does not see examines broad patterns of relationships between the social consumption of media messages and its stable, aggregate belief struc ( Shanahan and Morgan,1999, p.6).

PAGE 29

29 Recent cultivation theorists came to important conclusions about the relationship between crime and its media portrayal. Morgan, Shanahan, and Signorielly (2009) cite a few: Goidel, Freem an, and Procopio (2006) showed that attention to TV news in particular is associated with perceptions that juvenile crime is increasing. Watching reality crime shows is associated with the perception that overall crime is increasing. Holbrook and Hill (2005) found that viewing crime dramas was significantly rel ated to concerns about crime. Holbert, Shah, and Kwak (2004) examined relations between crime related TV and views on capital punishment and gun ownership. Eschholz, Chiricos, and Gertz (2 003) found that the perceived racial composition of neighborhood is a crucial dimension in structuring the TV/fear relationship, with television effects appearing primarily among individuals who perceive that they live in a neighborhood with high percentag e of blacks (p.43). Fear of crime is a pervasive consequence of the gap between the portrayal of crime in the media and the actual risk of being victimized. Despite the controversies about the direct effects of media exposure to crime on viewers, Dowler (2 003) claims that i ndividuals who frequently watch television are more likely to feel a greater threat from crime, believe that crime is more prevalent than statistics indicate, and take more precautions against crime Wood and Ribeiro (2011) point that: When reports of criminal activities are given headline status in the media and crime based plot lines become widespread, fear of crime and perceptions of its prevalence become relatively independent of the actual incidence of crime (Chiricos, Padgett and G uniform because the media make information on crime a daily occurrence, and the vicarious experience of victimization is emotionally stronger than Len and Zubillaga 2002:30). (n.p. ) In Latin America, media coverage of violence has several specificities and political consequences; most notably changing the way crime and justice are understood and with respect to instigating moral panics (Muraskin & Domash, 2007). In comparison to

PAGE 30

30 cr ime in the US media, Rey (2005) points out that cases like O.J Simpson, Robert Black or Michael Jackson are good examples of the relations between account, fiction, and cr imes are reported and many of them are never going to be punished. Impunity is 2005, p.12). Production studies analyze how producers relate to viewers and make decision s In production studies, the question of how producers relate to their audiences has been an important, if implicit, issue in understanding why producers make the decisions th at they do. Newsworthy studies of media producers living, working, and making decisions in their native habitats workers in the culture industry define as they go about making media con varying interpretations of who audience members are and what sort of imaginations of producers than in actual based merely on the best guesses of audience demand that producers can muster. (chapter 14, para.1) When it comes to crime shows, the process i s no different: producers claim they are just giving the audience what they want and the information they need to know. only define the central object and character of n ews stories, but are woven into methodology of journalists, influencing their choices from assignment, through the It is not within the scope of this study to determine the effec ts of violent news stories on people in Brazil. Instead, I will focus on what types of crime local media

PAGE 31

31 consider newsworthy, how is news produced, and how violence is framed on television, as well as the ways reports reinforce stereotypes, especially thos e of the victims. The importance of this study not only relates to the field of communications, but also has consequences for social sciences in general. The idea that crime and violence are widespread and out of control has severe consequences for societ y and the way it relates to its government and judicial system. In Latin America, Prillamam (2003) points that: The lack of confidence in the courts and police has prompted a growing number of Latin Americans to support quick rcuit or undercut democratic norms. The clearest manifestation of this trend has been the growing support for law and failings of the criminal justice system. (p.9) According to Wood and RIbeiro (2011), Th e sense of fear and vulnerability that taints Latin American cities has changed the ways in which people relate to one another, to the state, and to the very notion of citizenship (Rotke r 2002:17). Fea r of v iolence not only engulfs the excluded but also affects the wealthy and the powerful who contract private security details and barricade themselves in gate d communities (Caldeira 2000). (n.p.) Before turning to the specificities of violence on Brazili an media, it is important to understand how television achieved its supremacy as medium in the Brazilian society, and to identify the characteristics that shape the monopolies that own and control airwaves in the country. (Bucci, 1996). By the end of twenty one years of military dictatorship, Brazil was an urbanized countr y with one television set per 3.5 inhabitant and the influence of television in t he daily lives of Brazilians has grown since then. In 2008, 95,1% of the

PAGE 32

32 households in Brazil had a television set, being considered the second most important durable good in the home second only to the stove (PNAD 2009). In the midst of the 2008 econom ic crisis, Rede Globo the fourth biggest commercial network in the world registered the highest profits in its history: 7.5 billion reais. But t hese numbers only describe a fraction of what television represents in Brazil According to Bucci (1996): Wh at is invisible to the lenses of the TV does not belong to the Brazilian public space. It is within these [the television] limits that the country gets informed about itself, the world and recognizes its unit. In front of the screen, the Brazilians cheer t o gether, cry together during tragedies and laugh together. With a legislation that f orbids foreign capital in telecommunications, Brazil has five main networks, mostly family based companies. Rede Globo, founded by Roberto Marinho, is leader and has held a quasi monopol y in viewer ratings for more than three decades. According to Rockwell (2007): S uch television monopolies [in central and Latin America] or quasi monopolies may often represent strong antidemocratic forces in countries that are attempting d emocratic transitions after decades of war [In Brazil, de cades of Military Dictatorship] (p.35). M onopolies restrict the access to the media by several sectors of the society that are not related to the monopolistic group. The result is an impaired dialog ue, where the control of information is i n the hands of a few people whose interests ar e represented by the media. Monopolies held by television corporations neutralize other source s of popular culture dissemination. Alternative values to those associate d with the southeast urban/industrialized ones in Brazil are portrayed as exotic or marginal. (Sodr, 1984) The Brazilian television system is a public service and the companies hold temporary concessions under government control. Until 1988, the Execut ive had the

PAGE 33

33 power to grant concessions to d ifferent groups. Therefore, television growth since 1950 can be exp lained by political favoritism. The Executive power distributed licenses and concessions according to p olitical interests in accordance to the obj ectives established by the Nati onal Security Council. After the Constitution written in 1988, concessions must be approved by the National Congress. Most studies about the development of mass media in Brazil have identified the government as the main econ omic force behind the growth of the media by giving both economic and technica l incentives (Mattos, 2002 ). The advent of television in the 1950s happened during a perio d of intense industrial growth. In 1975, 60% of the Brazilian population already lived i n urban areas. (Mattos, 2002) The introduction of television in Brazil coincides with the beginning of changes in the economic, social and politica l structures in the country. By the early 60s there were already 15 television networks operating in Brazil. Mattos (2002) explai n s that external factors also influenced the development of the new medium of mass communication. In fact, during the 60s, both the United Nations and the United States advocated for the use of mass communication media to promote devel opment understood as economic growth and industrialization. During the military regime (1964 1985) the National Security Council determined achieved by the country. According to Selcher (in Mattos, 2002) the main national goals of that period were: national integration, sovereignty, development, progress and national prosperity, democracy, territorial integrity, and social peace. T elecommunications systems were an important actor in achieving these goals,

PAGE 34

34 promoting national integra tion and unity, preserving advocating for social peace (Mattos, 2002) In order to construct a new social order, the military regime needed a medium to tion media became the instrument used by the regime to persuade, im pose and spread its values. The television due to its immense potential for persuasion was the most used I n exchange, television networks benefited from t he infrastructure created for the development of t elecommunications. In 1968, television industry similarly benefited from a consumer credit that facilitated the purchase of products produced nationally. In 1968, television sales increased 48% (Sodre in Ma ttos, 2002). During the military regime, responsibility to foster development and promote national culture. The result, especially for Globo, was the reduction of imported shows which were replaced b y national pr oductions financed by nationa l banks. Currently, the networks try to limit the use of foreign productions to 30% of their programming (Mattos, 2002). While in the United States, the development of broadcast communications led to the hegemony o medium during an authoritarian regime led to the dominance of Rede Globo. In 2008, Globo registered an average of 24. 2% of the shares an impressive number, especially considering tha t it was the lowest one in the history of the network (IBOPE, 2008). According to Rockwell (2007): In the Brazilian context, the military in the 1960s fostered the growth of a dominant broadcast entity on television as a way to create a sense of

PAGE 35

35 nationalis m. The military government manipulated the advertising system to after the military era (p.47) About the 70s, Kehl (2004) explains: The 70s are strongly connected to the expansion of the c ultural industry in Brazil. To write about Brazilian television in this period is to reconstitute the history of the cultural industry in the country connected to the performance of the big economic monopolies, and consequently the history of Globo. The Br azilian television model was meant to be strongly liberal, although, in practice, it functioned to maintain government hegemony during the military regime, without any stimulus to competition (Simes & Mattos, 2005). Globo was created in 1965 under the C digo Brasileiro de Telecomunicaes (CBT), a legislation that establishes the primacy of the State in regulating the telecommunication systems. Simes and Mattos (2005) argue that the CBT never tried to balance the interests of the government, society and private sector, but instead was According to the authors: Mediating relations almost always clientelist ones [CBT] provided Globo the energy necessary to becom e the economic and political power that it represents today. In this environment, Globo developed; based on a regulation that has shown to be prone to the rise of hegemonic groups. ( Simes & Mattos, 2005, p.42) T he government invested heavily i n infrastruc ture and subsidized the production of television sets. Rede Globo was the main gro up that benefited from these policies (Santos & Capparelli, 2005). During its existence, Globo created and maintained a both from invested capital and te chnical knowledge. These barriers to entry into the television market in Brazil, thereby preventing competition. In

PAGE 36

36 other words, other media companies do not have access to the sa me capital that favored Globo during its first years and struggle d to invest in human capital and technology to meet the quality standards that Globo established Monopoly control of the most important medium of communications means the suppression of alte rnative voices concentration of political, economic and cultural p ower that Globo has prevent s competition Everything that is not o n TV and in this case, TV is understood as one dominant network is marginalized, leaving no space for public debate. Geographically distant from media centers, the state of Amazonas rarely makes it to the national broadcasts. As a consequence, local T V shows and newscasts are the only gatekeepers of information about what happens in the region. Currently, Amazonas media are concentrated in the capital, Manaus, and are dominated by four major family owned conglomerates that control television affiliates radio stations and newspapers. Among these conglomerates are Tv A Critica, that broadcasts Al Amazonas, and Band Amazonas, that broadcasts Comunidade Alerta.

PAGE 37

37 Figure 2 1. Media, process, and the perception of c rime. [ Barak,G.(1994) Media, process, a nd the social construction of crime (p.7). New York: Garland]

PAGE 38

38 CHAPTER 3 CRIME ON THE STREETS In recent years, the increase in crime and violence in Latin America has made it the most violent region in the world. With homicide rates comparable to war zo nes, and with El Salvador heading the list of most dangerous countries in the world, the region has been deeply affected by the rise in criminality The purpose of Chapte r 3 is to assemble data from various sources that can provide as objective a profile o f crime as possible: the incidence of different types of crime, the character of the crime (e.g., where it took place, what was stolen, etc), and how the types of crime vary by socio demographic characteristics, such as age, gender, income and so on. The r esult is a portrait of victimization and risk exposure that can further be compared to the characteristics of victimization in the media. With both portrayals in hand, I will then be able to empirically document the gap between the actual risk profile and the risk profile that is promoted on TV. To generate estimates of the prevalence of crime victimization, I used the 2009 National Household Survey PNAD. This questionnaire contains questions regarding to non violent property crime, violent property cr ime, and personal violent attacks. The goal is to provide a profile of who is more vulnerable to what type of crime. C hapter 3 is divided into lethal and non lethal victimizations. The latter include theft, robbery, and assault. Unfortunately, with the da ta provided by PNAD it is impossible to draw a profile of the criminal. For fatal victimizations, the national report Mapa da Violncia 2011 Os jovens do Brasil (Weiselfisz, 2011) provided the occurrence rate and some characteristics of homicides in the state.

PAGE 39

39 Before depicting the characteristics of crime, I generated frequencies to determine number of occurrences of robbery, theft and assault in relation to homicides (Table 1). The findings on Table 1 demonstrate that 99.77% of the crimes in the State of Amazonas fall into the non lethal category. Robbery (44.88%) is the most co mmon crime in the state, with 44.11 occurrences per 1,000 inhabitants, followed by theft (37.86 per 1,000 or 38.53%). Assault accounts for 16.36% of crimes, while homicides only account for 0.24%. Lethal V ictimizations Homicide is the most common ly used m easure of crime and violence in a society crime. Second, because homicide constitutes the final achievable point of a violent act against an individual, the most extreme d egree of violence possible. According to publicly visible of all violent acts and is usually reported more accurately in statistics on Since 1975, the Brazi lian Health Ministry has developed a Mortality Information System (SIM) that centralized all the information about the number and the causes of deaths in the country. The system follows the International Statistical Classification of Diseases and Related H ealth Problems ICD which accounts for external causes of the presence of an intentional assault of others, using any means to cause harm injury or death of (Waiselfisz, 2011 p.13 ) The system contains information about race/color with the same categories used by IBGE in PNAD, but the post mortem report is made

PAGE 40

40 by an external agent and does not, therefore, always match the self reported color/race of PNAD ( Waiselfisz, 20 11 ). Before starting the analysis of lethal victimizations in the state, it is important to point that analyzing homicide rates is not as straightforward as it seems. Wood and Ribeiro (2011) point to several problems that can lead to misinterpretation of d ata: First, the measure can be a misleading indicator of absolute violence because it does not include intentional but less than lethal injuries Second, the homicide rate for a country, by virtue of being a weighted average of the homicide rates that co rrespond to regional and local subunits, h ides within country disparities, which are often quite large. Within Latin America, cities are more violent places than rural areas, and, within cities, disparities in violence are associated w ith neighborhood inco me levels. The third caveat concerns the effects of changes in the quality and nature of reporting. Improved data collection can result in a statistical rise in violence that does not necessarily mean that violence has actually increased. Another pr oblem with homicide rate as a measure of violence is that it does not distinguish between types of homicide. Negligent homicide, self defense, involuntary manslaughter and murder, among others, are under the same category, despite the different levels of v iolence these incidents can represent. In Brazil, the Law makes a clear distinction between homicdio doloso (with the intent to kill) and homicdio culposo take int o account the intent or degree of violence used in the incident. Mello (1998) also questions the use of ICD to categorize the cause of death. where the doctor/medical autho rity claims it is impossible to know if the lesions were accidental or intentional. This category was supposed to be used only when investigation was not enough to determine if the case was an accident, suicide or homicide. However, Mello (1998) notes:

PAGE 41

41 The excessive number of existing cases under this category, in some areas, is due to the fact forensic professionals rely on an ethical behavior in which their obligation is only to describe the injury (medical cause of death) and not determine what kind of a ccident or violence caused the same injuries (legal cause of death). Ideally, to make general inferences about the level of violence on the basis of homicide rates requires going beyond counting bodies and classifying injuries. It requires collecting infor mation about the incident as a whole. Unfortunately, the lack of forensics data and the inability to link existing police data and SIM makes it impossible to provide a full profile of homicide incidents, putting together information about the criminal, the incident and the victim. To grasp some characteristics of lethal victimizations in the state of Amazonas, I used the national report Mapa da Violncia 2011 Os jovens do Brasil (Weiselfisz, 2011). This report provides an overview of violent deaths in Br azil from 1998 to 2008, focusing on youth. The report treats car/motorcycle accidents, suicides and homicides as violent deaths, defined as e the result of some action of indivi duals, either against himself a s in the case of suicide, or by the intervention, whether intentional or not, of other people (Weiselfisz, 2011, p.10). For the purpose of this study, other than general homicide rates, the report provides youth homicide rates and racial profile of victims from 1998 to 2008. In Brazil, the North region where Amazonas is located the 1998 2008 period witnessed the highest growth of absolute homicide numbers (+108. 1 %) and the third highest increase in terms of homicide rates per 100,000 (+63.1%). However, this rate is very unequal across the seven states in the region. In 2008, the state of Amazonas had 827 homicides, an increase of 54.3% in a decade, making the state the 17 th most dangerous state in absolute terms. By

PAGE 42

42 comparison, Par the most populous state of the North region had 2,868 homicides in the period, an increase of 273% in the last ten years. Rio de Janeiro had 5,395 homicides ( 28.7% since 1998), and So Paulo had the highest number of homicides in the year (6,118) despite a sharp decre ase of 56 3 % in the number of occurrences in the last decade. In regards to homicide rates per 100.000 inhabitants, Amazonas has a rate of 24.8 in 2008, an increase of 16.4% in a decade. In comparison to other states, Par had a rate of 39.2 ( +193.8%), Rio de Janeiro had a rate of 34 ( 38.6%) and Sao Paulo had a rate of 14 9 ( 62 4% ). Alagoas is the state with the highest homicide rate in Brazil, with 60.3 occurrences per 100,000 inhabitants, which was 177.3% higher than the previous decade. The capit al of Amazonas, Manaus, was the location of 656 (79.3%) of the homicides in the state with a rate of 38. 4 homicides per 100,000. Although Manaus has a rate higher than the state as a whole, this rate actually decreased ( 5.6%) in the last decade, while th e state rates increased by 16.4%. Noting the changes over time, Waiselfisz ( 2011 ) called attention to the increases that have taken place outside of metropolitan areas, usually in the small towns and rural places, called the interior. According to the aut hor: This differential rhythm with metropolitan regions and capital stagnating or falling while the interior continues to grow is what we called, ever since 2002, internalization of violence indicating a change in the dynamic poles. This does not mean t hat the rates of the interior are larger than those of large urban conglomerates. It simply means it is the interior that is responsible for the growth rates of homicide, and no longer the capital or major cities. (Waiselfisz, 2011, p.51)

PAGE 43

43 The report also e mphasizes the homicide rates among the youth. Among people 15 24 years old, Amazonas had 319 homicides (46 occurrences per 100,000 inhabitants). That means 47.1% of the total homicides in the state happened in this age group alone. Young people are nearl y twice as likely to be murdered compared to the overall population. When reporting homicides by race Mapa da Violncia called negro the sum of people in the categories black and brown/mixed ethnicity used by IBGE. I will call this category Afro Brazilia ns. Once again, there are some disparities between the SIM data reported by doctors or medical authorities and IBGE self declared race data used in PNAD. In 2008, the homicide rate among Afro Brazilians was 29.5 per 100,000. This rate was 4.3 among whi tes in the state. Waiselfisz (2011) defined the difference between the white homicide rate and the afro rate was 580.6 in Amazonas. That means that the probability of a victim of homicide being Af ro Brazilian is 580.6% higher than being white. Homicide data, although more straightforward, has many disadvantages: homicides are a rare occurrence, hence most people are not affected by each incident. More commonly, people are afflicted by robbery, thef t and assault. Whereas official records of these more common crimes exist, the statistics a woefully inadequate. Therefore, to have a portrayal of violence and crime in the streets of Amazonas, we turn to self reported victimization surveys, the most compr ehensive of which is the Brazilian National Household Survey (Pesq uisa Nacional por Amostragem Do miciliar, PNAD).

PAGE 44

44 Non lethal V ictimizations Every year, the PNAD investigates the general characteristics of the Brazilian population, including education, work income and housing from all of the twenty six states of Brazil and the Federal District (Distrito Federal). In 2009, PNAD interviewed a total of 399,000 respondents For that year, PNAD contained a supplement with twenty four questions regarding victimi z ation. With this dataset, it was possible to measure the occurrence of nonfatal crimes reported and non reported in the state of Amazonas. Because PNAD is based on survey with victims, it is impossible to measure homicide s Using cross tabulations, I obta ined characteristics of victimization occurrences in the state. The questionnaire contained questions including characteristics of the occurrence such as place of crime, what was robbed/stolen, relationship with the aggressor, etc. d emographic information. PNAD 2009 provides information on three types of crime: theft, robbery, and assault. With this dataset, it is possible to answer the question of who is victimized and where it happens and what was involved. Although providing detai led measures of crime and its characteristics, Wood and Ribeiro (2011) identified some weaknesses of victimization surveys: Targeted surveys of this sort have the potential to generate accurate estimates of crime victimization rates, but they suffer from a t least three drawbacks. First, surveys can be subject to sampling error such that the sample is not representative of the popul ation from which it is draw. Second, because most victimization surveys are based on a small number of cases, they are not representative of both national and subnational units. Third, dedicated victimization surveys, unlike the self consciously comparative surveys such as the International Crime Victimization Survey (ICVS), are rarely comparable across countries. The third p roblem is evident when dealing with PNAD data. The self reported questionnaire does not make a distinction between crimes against property and against

PAGE 45

45 personal crime, nor does it explicitly account for rape or sexual assault. When compared to the United St used a much simpler questionnaire. The US report includes detailed categories such as sexual assault; aggravated/simple assault; household burglary; and property theft. These categories are divided in to the following groups: non violent crime, violent crime, serious violent crime, personal theft, and property crimes. The lack of some of these distinctions is a limitation of PNAD 2009. For example, when a person reports he/she was a victim of robbery in a commercial establishment, it is impossible to establish if the victim was the owner of an establishment that got robbed, the cashier who handled the money from the establishment, or a costumer that got robbed when shopping in that location. In PNAD, e ach respondent could have been the victim of more than one time during the year. By aggregating those numbers I estimated the total number of occurrences of crime. The sample was weighted by the individual sampling unit (V4729) such that sample accurately depicts the population. The results show the characteristics of occurrences and characteristics of victims of such crimes. General Risk Overall, the most common type of non lethal victimization in Amazonas is robbery, followed by theft and assault (Table 3 1). The definition of robbery by PNAD is theft with the use of violence. It is important to note this variable includes personal and property robbery, and makes no distinction between the armed and no armed occurrences. Residential and commercial establ ishment burglary, without threatening the physical integrity of any individual, is under the category of theft If the burglary included using

PAGE 46

46 violence or threatening to use violence against an individual, then it falls into the category of robbery. A phys property is considered assault. Non lethal victimization varies by location, most targeted items, and closeness to the criminal Table 3 2 describes the characteristics of victimization occurrence s in the state of Amazonas from September, 2008 to September, 2009 according to self reported data These characteristics include: items stolen/robbed; place of crime; and relationship between victim and assailant. Cell phones are the most targeted items for theft and robbery, more than money/credit card. Most violent crimes occur in via pblica which means a public location such as a street or a sidewalk, but can also be a park or a bus stop, for example. A clear distinction between the profile of victims of property crimes and violent crimes is observed. According to Buvinic and Morrison (in Wood & Ribeiro, 2011), ore prosperous areas suffer from violent property related crimes, and severe violence is generally con centrated in lower Brazilians are more than four times more vulnerable to assault (Table 3) and six times more vulnerable to homicides than whites. Younger populations are also more likely to be victimized by violent crime. As of homicides, youth homicide rates are twice the overall population rates. Table 3 3 contains the characteristics of the victims of theft, robbery and assault in the state. With T able 3 3 it is possible to infer who are the most exposed to the risk of being victimized by age, sex, location, race, employment status, occupation, and

PAGE 47

47 income. The most vulnerable age group for non lethal victimizations includes people from 40 49 year s old (163.48 per 1,000 inhabitants). The racial profile of the victims shows that blacks the most vulnerable group to all types of non lethal crimes in the state with 164.9 victimizations per 1,000 inhabitants. Whites are the second most vulnerable (101), followed by brown/mixed ethnicity (96.6). There were no incidents reported by indigenous or yellow (Asians). It is worth mentioning that brown/mixed color group accounts for 77% of the state populati on, followed by whites (21%)( Table 3 4 ) While blacks are only 2% of the overall population, the victimizatio n rates among the black group are 1.6 times higher than whites and 1.7 higher than brown/mixed ethnicity. To estimate the distribution of victimization by social class, I used household income as operational definition. PNAD measures household income b y minimum wages earned a month I n 2009, the minimum wage was 465. 00 reais. This is the equivalent of almost 260 dollars per month, or an annual income of a little less than 3,118 dollars 1 In general people whose household income is higher (b etween 5 to 10 minimum wages) are the mos t vulnerable to become victimized with a rate of 147.09 occurrences per 1,000 people (Table 3 3 ). In regards to employment, the most vulnerable group to non lethal victimizations Among this group, the rate is 145.95 occurrences per 1,000 inhabitants, while the unemployed had a rate of 98.73 per 1,000. 1 Using the dollar rate of 08/26/2009. $1 USD = R$ 1,79

PAGE 48

4 8 PNAD classifies occupational groups into nine broad categories: general managers; science and arts; technicians with high schoo l degree only; administrators; services in general; sales and commerce; agriculture; construction, repairing and maintenance; and armed forces and police. There is a tenth group called undefined occupation, which had zero respondents. The occupational gro up that most suffered more victimization in 2008/09 were the sales and commerce workers, with a rate of 230.97 occurrences per 1,000 people. The second most vulnerable group included workers in construction and in repair and maintenance services (193.46 p er 1,000). Agriculture workers were the least vulnerable target of crime (25.97 occurrences per 1,000), with a rate of zero robberies and very low assault (3.3 per 1,000) reported during the year. These data confirm the urban character of violent crimes in the state: non lethal crimes in Amazonas are predominantly urban, with rates more than nine times higher than rural areas. Theft A theft threat of use of violence. For this categor y, respondents could report more than one item; therefore, in the same occurrence, it is possible that money, cell phone, documents and personal items were stolen, for example. The most common items stolen were cell phones (37.09% of the times), followed b y money (26.77%) and documents/personal items (21.95%). The theft of vehicles accoun ted for 4.84% of the occurrences (Table 3 2). cations. Public locations, such as streets,

PAGE 49

49 sidewalks and bus stops, are the second most vulnerable places to theft (26.70%). Public transportation is the third (12.8%) followed by commercial establishments (8%) This accounts for people stealing one item from a convenience store to thieves burglarizing a department store (Table 3 2). Most of thefts occur in an urban environment, with a rate of 44.58 occurrences per 1,000 inhabitants in cities against 4.91 per 1,000 in rural areas (Table 3 3). Theft is the only type of crime for which females are more vulnerable than males 38.9 per 1,000 versus 36.82 of their male counterparts. Blacks are more likely to become victims of theft (55.5 per 1,000), followed by whites (43.09 per 1,000) and browns (36.17 per 1, 000). Employed people are almost twice more likely to be stolen, with sales and commerce being the most targeted occupational group (99.94 per 1,000). People with an income between 3 to 5 minimum wages are the most exposed (70.34 per 1,000). The age group of 40 49 years old is the most targeted (72.91 per 1,000), followed by people between 50 and 59 years old (Table 3 3 ). These numbers show that thefts are the most likely to occur in public locations, aiming to steal cell phones. The most exposed people are black middle class adults, especially those working in sales and commerce. While males are more exposed to violent crimes, females are slightly more subject to theft.

PAGE 50

50 R obbery Robbery 2 is the most frequent type of crime in the state of Amazonas. 44.11% of use or threat of use of violence (Table 3 1). Cell phones are the most frequently robbed item in the state (38.68%), followed by money (31.06%) and documents/persona l items (15.53%). The riskiest places for being robbed are public locations (73%), such as streets or sidewalks. The second most vulnerable place ( Table 3 2). The difference between robbery rates in urban and rural areas is remarkable: 52.37 victimizations per 1,000 inhabitants in urban areas versus 3.6 in rural ones. The higher income group (10+ minimum wages) is the most vulnerable to robbery, with rates of 60.97 occurrences per 1,000. (Table 3 3) Robbery occurs more common ly among males (46.78 per 1,000 versus 41.48 per 1,000 female victims) and those who are employed (66.39 per 1,000). Blacks are the most vulnerable race to robbery (87.6 per 1,000), followed by whites (48.03) and browns (42.22). The most vulnerable age g roup is 30 39 years old (69.03 per 1,000), followed by 40 49 years old (66.03 per 1,000). In regards to occupational groups, the most victimized are those who work with sales and commerce (113.5 per 1,000), followed by those who work with repairing and ma intenance (94.74 per 1,000) (Table 3 3). 2 According to the definition by PNAD questionnaire robbery includes mugging, property burglary if residents are threatened, with or without weapons. These definitions vary from country to country, making it hard to make cross country comparisons (Wood & Ribeiro, 2011).

PAGE 51

51 Assault Assault is the only non property related crime listed in PNAD questionnaire. It includes any type of physical aggression, but the question is not specific about sexual assault/rape or verbal aggression, whic h are included in the definition of assault in the United States. There are no sub categories of assault, such as aggravated/simple, sexual/rape, etc. Also, PNAD does not distinguish attempted homicide from assault. In n to kill is not taken into account. In regards to assault, the most vulnerable places are public locations. 58.6% of the assaults in Amazonas happen on the streets, roads, sidewalks, etc. (Table 3 2). The residence (20%). About the relationship with the assailant, most victims are assaulted by unknown people (45.45%), followed by acquaintances (35.5%). Assaults by a partner/former partner account for 9% of the occurrences. Overall, these numbers show that a ssaults are mostly random and public committed by strangers in public locations (Table 3 2). Younger populations (20 29 years old) are the most vulnerable with a rate of 25.45 per 1,000 inhabitants, followed by people from 10 19 years old (24.8 per 1,00 0). The gap between males and females is higher than other non lethal victimizations: 20.25 assaults per 1,000 for males and 11.7 per 1,000 for females (Table 3 3). Compared to whites, blacks are twice as likely to be assaulted. If we put together blacks and browns in one category, whites are more than four times less vulnerable to assault than afro Brazilians in the state of Amazonas (Table 3 3). Assault is also the only type of non lethal crime that is more common aiming the unemployed: 21.65 occurrences per 1,000 against 18.24 on the employed group. With

PAGE 52

52 respect to occupation, high school level technicians were the most victimized, with a rate of 41.96 per 1,000. Lower income people are at the highest risk of assault. The higher rate falls into the cate gory between half and one minimum wages (23.39 per 1,000 inhabitants). There is a clear distinction between the vulnerability to crimes that involve property (theft and robbery) and those that involve physic al aggression (assault). R obbery is considered a violent crime, as it involves the use or threat of use of force When it comes to assault, a crime involving violent physical attack without the purpose of stealing some property, the lower income young dark skinned citizens are the most vulnerable. It i s worth highlighting that assault is also the type of crime that is more r andom in the majority of cases the assail ant is a stranger. Overall, the most vulnerable people to become vict imized by a non lethal crime live predominantly in an urban area, wor k in sales and commerce, and are black adults. However, is it important to note that the more violent the crime, th e younger, poorer, and darker the victim. (Figure 3 3, Figure 3 4, Figure 3 5 )

PAGE 53

53 Table 3 1. Victimization rates per I,000 inhabitants Amazo nas, Brazil 2008 /09 Type of Crime Rate Percentage Theft 37.86 38.52% Robbery 44.11 44.88% Assault 16.08 16.36% Total Non lethal V ictimizations 98.06 99.77% Homicides* 24.5 0.24% Total V ictimizations 100% *Per 100.000 inhabitants. Source s : Mapa da Violencia 2011 and PNAD 2009. Table 3 2 Victimization rates per I,000 inhabitants Characteristics of the victimizations, Amazonas, Brazil 2008 /09 Crime Description Rate n Percentage Theft 37.86 What was stolen? Money 26888 26.8 Credit/Debit card 8748 8.7 Cel l Phone 37253 37.1 Documents/Personal Items 22047 21.9 Vehicles 4858 4.8 Other 648 0.6 total 100442 100.0 Where did the theft happen? Own residence/ others'residence 40760 46.0 Commercial Establishment 7126 8.0 Public location 23648 26.7 School/Educational institution 2916 3.3 Public transportation 11338 12.8 Stadium or gymnasium 647 0.7 Other 2267 2.6 Total 31915 100.0 Robbery 4 4.11 What was robbed? Money 60765 31.1 Credit/Debit card 13282 6.8 Cel Phone 75661 38.7 Documents/Personal Items 30378 15.5 Vehicles 3887 2.0 Other 11637 5.9 T otal 195610 100.0

PAGE 54

54 Table 3 2 Continued Crime Description Rate n Percentage Robbery (Continued) Where did the robbery happen? Own residence/others' residence 13907 12.4 Commercial Establishment 4858 4.3 Public location 82141 73.2 School/Educational instit ution 1942 1.7 Public transportation 8099 7.2 Stadium or gymnasium 0 0.0 Other 1295 1.2 Total 112242 100.0 Assault 16.08 Relationship with ag g ressor Unknown person 14483 45.4 Police 952 3.0 Private security 324 1.0 Partner/Former partner 2894 9.1 Relative 1923 6.0 Acquaintance 11339 35.5 Total 31915 100.0 Where did the assault happened? Own residence 6436 20.2 Others' residence 950 3.0 Commercial Establishmen t 1295 4.1 Public location 18718 58.6 School/Educational institution 972 3.0 Public transportation 324 1.0 Stadium or gymnasium 1620 5.1 Other 1600 5.0 Total 31915 100.0 Source: PNAD 2009

PAGE 55

55 Table 3 3. Victimization rates per 1,000 i nhab itants Characteristics of the victims, Amazonas, Brazil 2008/09* Theft Robbery Assault Total Age 10 19 22.6 38.55 24.8 85.96 20 29 48.35 67.02 25.45 140.82 30 39 59.53 69.03 16.22 144.78 40 49 72.91 66.03 24.54 163.48 50 59 67. 33 48.17 11.4 126.9 60+ 32.67 35.96 2.7 71.42 Sex Male 36.82 46.78 20.52 104.12 Female 38.9 41.48 11.7 92.08 Place of R esidence Rural 4.91 3.6 3.61 12.12 Urban 44.58 52.37 18.63 115.58 Race Indian 0 0 0 0 White 43.09 48.03 9. 8 101 Black 55.5 87.6 20.97 164.9 Yellow 0 0 0 0 Brown 36.17 42.22 17.69 96.6 Employed** Yes 60.72 66.39 18.84 145.95 No 33.45 43.63 21.65 98.73 Occupation** General Management 91.73 36.4 18.21 146.34 Science and Arts 51.32 19.65 0 .00 70.98 Technician 41.93 67.78 41.96 151.7 Administrative positions 71.48 65.96 2.70 140.19 Services worker 48.49 73.81 31.10 153.41 Sales and commerce 99.94 113.5 17.49 230.97 Agriculture 22.59 0 3.30 25.97 Repairing and maintanance 75.25 94 .74 23.46 193.46 Armed Forces and Police 28.99 28.99 14.49 72.47 Income** < 1/2 Minimum Wage 11.82 18.66 4.97 35.46 More than 1/2 to 1 32.21 36.21 23.39 91.82

PAGE 56

56 Table 3 3. Continued. Theft Robbery Assault Total More than 1 to 2 37.61 50.1 16 .15 103.87 More than 2 to 3 58.87 61.1 8.88 128.86 More than 3 to 5 70.34 42.95 11.19 124.49 More than 5 to 10 65.83 65.76 15.49 147.09 10 Minimum Wages + 43.06 69.97 16.14 129.18 Source: PNAD 2009 *p<.001 **Defined by having worked last week **Valid percent of total population 47%. Excludes missing data. Table 3 4 Racial composition Amazonas, Brazil 2008 /09 White 721568 Black 58351 Brown 2667276 Yellow 2592 Indian 4774

PAGE 57

57 Figure 3 1. Racial composition Amazonas, Brazil, 2008 /09. Source: PNAD 2009 Figure 3 2. Non lethal occurrences by race (rate per 1,000 inhabitants) Amazonas, Brazil, 2008/09 21% 2% 77% 0% 0% White Black Brown Yellow Indian 0 20 40 60 80 100 120 140 160 180 Theft Robbery Assault total

PAGE 58

58 Figure 3 3. Non lethal occurrences by age (rate per 1,000 inhabitants) Amazonas, Brazil, 2008/09 Figure 3 4. Non lethal occurrences by income (rate per 1,000 inhabitants) Amazonas, Brazil, 2008/09 0 20 40 60 80 100 120 140 160 180 10-19 20-29 30-39 40-49 50-59 60+ Theft Assault Robbery Total 0 20 40 60 80 100 120 140 160 < 1/2 Minimum Wage More than 1/2 to 1 More than 1 to 2 More than 2 to 3 More than 3 to 5 More than 5 to 10 10 Minimum Wages + Theft Robbery Assault Total

PAGE 59

59 CHAPTER 4 CRIME ON TV To document the way media construct public images of the prevalence and character of crime, this section analyzes stories from Comunidade Alerta and Al A mazonas between June 06 17, 2011. I used a data collection protocol developed by Cavender and Bond Maupin (1993) in their analysis of crime shows in the United States. The protocol focuses on three aspects of programming: demographics (general informatio n on criminals and victims); characterizations (brutality, dangerousness, view (relative safety of people, places and attitudes the audience should have) (Cavender & Bond Maupin, 1993). To address the gap between crime a s a media discourse and reality, I used interreality 1 comparison measures. Dixon and Linz (2000) developed this measure in their study of race and misrepresentation of victimization in local news programs. Although other types of comparisons, such as inter group (e.g. black criminals and black victims) and interrole (e.g. black victims and white victims) comparisons, can offer valuable insights of the portrayal of groups on the news, it is the interreality measure that offers information pe rtinent to my rese arch hypothese s. According to Dixon and Linz (2000): Interrole and intergroup comparisons allow us to measure the television social world after exposure to television. However, th ese comparisons tell beliefs that are subsequently cultivated in viewers. It may be the case that Blacks are, in fact, more likely than Whites to commit crime depending on what i ndex of crime is used as a comparison point. In this sense, television news may disseminate an accurate picture of the world rather than a 1 The most recent data available from PNAD and Mapa da Violncia date from 2008. None of the TV stations had an archive with the shows during this time frame.

PAGE 60

60 distorted view. Intergroup and interrole measures must therefore be anchored to objective indicators of social realit y that are measured outside the television environment, both in order to evaluate claims of the accuracy of the media representations and in order to provide a basis for speculating and evaluating the potential effects on viewers (p.135). Sampling This stu dy uses convenience samples from Al Amazonas and Comunidade Alerta. Unfortunately, Band Amazonas does not keep an archive of previous programs, which made it impossible to get a systematic random sample of Comunidade Alerta. They have some old recorded ep isodes that were used for coder training. TV A Crtica has a reasonably organized archive, which contains all Al Amazonas episodes from the last few years. I used all the dates when both shows were available. The 20 episodes analyzed contain 58 stories 42 from Al Amazonas and 16 from Comunidade Alerta. Despite the fact that these shows were already broadcasted, access to the episodes of the sample was one of the most challenging parts of the research. Constantly facing lawsuits, the stations were not w illing to provide materials that could be used against them later. By adding ethnographic information on news making process, I sought to minimize the limitations of the sample, as well as not limit my research to a content based analysis. This study did n ot focus on the portrayed characteristics of the criminals in the shows, as it was impossible to proceed with an interreality comparison using victimization data. A convenience sample has limitations, as it is not representative of the whole content broadc asted, and therefore generalizations cannot be made. Nevertheless, Riffe, Lacy and Fico (2005) explain that convenience sample can be justified in three cases: when material is hard to obtain; when resources for a systematic random sample

PAGE 61

61 are limited; and when researcher is exploring an underresearched topic. This study falls into the three categories, which is not uncommon when conducting content analysis on local television content. According to Riffe, Lacy and Fico (2005): The main reason nonprobability samples are used is difficulty in collecting content. As a result, the availability of content shapes the nature of the research. The study of local television news provides a good example. Local TV newscasts are rarely available outside the particular mar ket. As a result, the programs must be taped in the area of origin. (p.101). I used all the episodes available (census), which resulted in a different number of stories per show. Comunidade Alerta broadcasts an average of two stories per episode, intersper average of six smaller stories per episode, dedicating the last segment to commentary and opinionated pieces. Coder Training and R eliability Three Portuguese native speakers, includi ng myself, were used in reliability tests. They were graduate students in Manaus, one male and two females. Training took place over three days and each section lasted 2 hours. During these sections, the coders identified and helped solve coding problems w ith the codebook. For the training, website, and were not part of the final sample. For the reliability tests, each coder received 6 randomly selected identical programs (25 stories 9 from Comunidade Alerta and 16 from Al Amazonas, 43% of the total sample) that provided the reliability data for this study. The episodes were randomly chosen from the final sample. After reliability was achieved, the researcher did the coding

PAGE 62

62 coefficients ranged from 0.74 to 1.00. I eliminated two variables that did not attain desired reliability 2 Each coefficient is reported in the following sections. Unit of A naly sis The units of analysis were crime stories from TV shows Comunidade Alerta and Comun idade Alerta spend a considerable amount of time talking about politics and defending themselves from attacks regarding their electoral charges. These minutes were not coded. Crime stories were considered those where an offense was depicted. All other stor ies were not coded. The coders analyzed the images displayed as well as the discourses presented. Coding Main C rime The first thing coders identified was the main offense depicted in the story. It is important to note that crime and violence are not syno nyms. Not all crime is violent (e.g. corruption, theft), and violent acts might be considered legal in some Latin American countries (e.g. state violence, domestic violence) (Prillaman, 2003). The coders were instructed to code all stories about crime, inc luding non violent ones. There were no stories about violent non criminal acts depicted, but communication researchers should be aware of this instance when coding news stories in Latin America. 2 a) From a scale of 1 to 7 (1=negative, 4= neutral, 7=positive), how are public policies portrayed in the story? b portrayed in the story?

PAGE 63

63 A few stories depicted more than one offense. For example, ro bbery followed by homicide, or drug trafficking and gang organization. In those cases, only the main crime depicted was coded, meaning the one that the story emphasized the most. Reliability for main crime depicted was = .87. When the main crime was a l ethal victimization, coders also coded for motive ( = .85). Coding V ictims The coders first needed to determine if victims and offenders were depicted. Although this variable seems straightforward, it was actually hard to define what could be considered t he depiction of a victim. Body parts, corpses and body bags were coded separately from the victim alive in case of non lethal victimizations, or prior to lethal victimization in case of homicides (pictures, videos). After making this distinction, the coder s reached perfect agreement on these categories ( =1). In regards to how is the victim portrayed, coders had a scale of 1 to 7 (1=negative, 4=neutral, 7=positive). The reliability for this variable was 0.87. Codin g Demographic C haracteristics gender, age, race, place of victimization (rural/urban) and occupation race, yellow (Asians), and Indians/native Brazilians. Reliability was sufficient ( =.76) using five c ategories. However, since Mapa da Violencia uses the categories whites and afro Brazilians, I also summed blacks and browns into a broader category called afro Brazilians. There were no cases in the sample in which Asians or Indians were depicted as victim s. Using two categories (whites and afro Brazilians), the coders reached perfect agreement ( =1).

PAGE 64

64 Age was coded according to what the story explicitly mentioned ( =.96). The same approach was used to code occupation ( =0.96) and place of victimization ( =1 ). =1). If the coders could not determine the gender and race, or if the age, occupation and place urther research about any victimization was conducted in newspapers or police records. Results Television depictions of crime in the state of Amazonas differ significantly from PNAD official statistics on victim ization. The findings in Table 4 1 show that the most common offense depicted on the shows was homicide. Lethal victimizations represented 62.5% of the crimes depicted in Comunidade Alerta and 33.3% of the stories from Al Amazonas. While homicides represent only 0.24% of the actual victimizations i n the state of Amazonas, it is the most newsworthy occurrence and accounts for 41.4% of the crime stories broadcasted by the shows combined. For most homicides depicted (36.7%) the motive is unknown. Money (12.5%) and no apparent reason (12.5%) are also c relates to those crimes where the story explicitly says that the victimization happened host/commentator actually reinforced the ide falta de Deus no corao) or because the murdered wanted to do evil ( querendo fazer o mal). Half of the homicides depicted by Comunidade Alerta did not depict any motive. The rest of the homicides attribu ted the crime to drugs disputes or addiction.

PAGE 65

65 To test H ypothesis 1, I created an index of depiction of violent elements, by combining information on images of drugs, weapons, ammunition, blood, corpses, wounds, body parts, and/or body bags. This index allo wed me to compare the display of violent images between the shows (T able 4 2 ). The findings show that Comunidade Alerta broadcasted an average of 2.18 violent elements per story. A homicide story, for example, depicted the weapon of crime and blood; or a drug burst depicted the drugs and ammunition seized. Al Amazonas used an average of 1.19 elements per story. An element could not be coded twice in the same story. The differences between t he main crimes depicted (T able 4 2 ) also demonstrate that Comunid ade Alerta is more violent than Al Amazonas. The political show only broadcasted stories about homicide and drug disputes, while Al Amazonas had a more diversified coverage. In fact, the most common offense depicted by Al Amazonas was robbery (35.7%), w hich is the most common crime according to official statistics. Although homicides were overrepresented in both shows, Comunidade Alerta limited its coverage to violent lethal victimizations and drug related stories. The show also used twice as many viole nt eleme nts on its stories. Therefore, H ypothesis 1 is accepted. With respect to the relationship between the show and politics, the Comunidade Alerta viewers I interviewed did not see a problem with the fact that the Tabosas were politicians. S., 50 year men that show what they are doing. And we can only vote for those who can prove they years

PAGE 66

66 cook, agrees. She beli eves that voting for traditional politicians has proven to be so disappointing that she prefers to vote for the hosts, who at least show they are doing their jobs on TV. When I questioned them if the 2004 scandals have affected crime a lot of good things. Even with all their mistakes, they helped a lot of people, killed a lot communicatio n, J une 17th, 2011). The viewers interviewed say that they know the hosts use the shows for political gain, but they still prefer to vote for them. J., 48 years old, do something directly. Changing this it will take time. These politicians are aware of we vote this way, and they do stories in poor neighborhoods, where the electorat (personal communication, J une 17th, 2011). In regards to the way victims are portr ayed in crime shows, the analyses of Comunidade Alerta and Al Amazonas revealed that victims (excluding portrayals of the deceased) appear on only 41.4 % of the stories analyzed. However, the chances that the body of a victim will appear are higher (52.4% ) ( T able 4 3 ). When the victims are portrayed, 77.7% of them are men, with the majority falling into the 18 29 age group. These findings are in accordance to Mapa da Violencia data that show that homicides are more prevalent among younger males. The find ings i n Table 4 4 suggest that the portrayal of non lethal victims is not in accordance with PNAD victimization rates. First of all, there is a significant underrepresentation of female victimization, with only 17.4% of the stories portraying female victim s of theft, assault, or robbery.

PAGE 67

67 In regards to age group, the TV shows only showed significantly younger victims, all of them below the age of 29 years old. The most victimized age group in PNAD is 40 49 years old, followed by 30 39 years old. Brown/mixe d race is the most victimized race on television. In the aggregate, this is an accurate depiction of reality simply because browns far outweigh the other categories in the racial composition of Amazonas. However, in comparing the victimization rates of the different races to media portrayals of victims, a different picture emerges. These results show an underrepresentation of black and white victimization. When comparing the portrayal of non lethal victims and PNAD data, the white and black underrepresentat ion becomes more evident. Blacks are portrayed as victims on 4.3% of the non lethal stories, while they are the most victimized race in PNAD, with a rate of 164.9 victimizations per 1,000 inhabitants. Whites were portrayed as the victims in only 8.7% of th e non homicide stories, whereas they are the second most v ictimized group in PNAD (Table 3 3 ), with a rate of 101 victimizations per 1,000 inhabitants. This means that the portrayal of victims in the TV shows gives the impression that browns are at the m ost risk of being victimized, when in fact, the opposite is more accurate. While in absolute terms there are more brown victims in the state, the rate among the browns is the actually the lowest (96.6 per 1,000). There were no reported crimes that involved Indians or Asians in the period analyzed. Both shows had similar patterns in regards to gender and race. Overall, the victims tend to be neutrally portrayed. In a 1 7 scale, the average portrayal of victims is 4.4. Whites and non depicted victims are por trayed more

PAGE 68

68 positively t han afro Brazilians ( Table 4 5 ), and males are portrayed more negatively than females (Table 4 6 ). Generally, victims tend to be portrayed more positively in the media, yet intergroup comparisons between whites and non white victims are more likely to be identified in negative contexts (e.g. crime), even when they are overrepresented as victims on local TV shows, but a lso negatively represented when compared to white victims. Female victims, on the other hand, tend to be depicted in negative roles in the North American media (Barak, 1994), while programs in Brazil, females were underrepresented, but positively portraye d in comparison to their males counterparts. It is worth noting that, although the results are statistically significant, the underrepresentation of whites (4 cases) and females (6 cases) affects the strength of these findings. The most victimized occupat ional groups on television are people working with sales and commerce (13.3%) and students (13.3%). While commerce is also the most victimized occupational category in PNAD data, being a student is not considered an occupation by the victimization survey. The portrayal of commerce victims comes mainly from stories about robberies that happened in a commercial establishment. Although the portrayed characteristics of the victims show interesting results, it is stories that has proven to be the most

PAGE 69

69 same stories the occupation is omitted; and in 51.1% the race is not shown or described. The victim does not appear at all in 36.2% of the stories. In the shows analyzed, victims of robbery almost never showed their faces, claiming fear of retaliation, although the location of the robbed establishment and full reenactment of the crime scene were considered safe. Caldeira (2000) explains that, in Brazil, an act of crime serves as a turning point for the victim. According to the author: For the sake of rhetoric, the division between before and after [the victimization] reduces the world and the experience. The before becomes too go od ; the after becomes too bad. Descriptions of previous happiness are romanticized: the house with the marble staircase, swimming pool, and barbecue; the diamonds worn on an ordinary afternoon; the relaxing interludes at the piano; in short, comfort, order, and status, all interrupted by the fateful doorbell. After the assault [described in a previous interview], life is hell: everything has lost its savor, she [interviewee] and her husband have lost their health, her son is full of fear, they have lo st money and status. (Caldeira, 2000, p.28) In the case of Amazonas, victims of robbery and theft were often the ones who called reporters from Al Amazonas and Comunidade Alerta to cover the story, even though they would decline to be interviewed. Instead their goal was to advertise the were ashamed of their status as victims. Fabiola Gadelha, reporter from Al Amazonas, s unwillingness to collaborate, although sometimes she convinced them to speak by promising that their identities would be preserved. About the lack of focus on victims and their characteristics, Katz (1987) states: A focus on victims would have substanti al practical relevance for readers, were they concerned with learning h ow to avoid the costs of crime. the emphasis of the city papers on criminal rather than the victim is avoiding vic timization than in working out moral positions on which his own behavior will be based. (p.235)

PAGE 70

70 When I questioned viewers as to why they watch crime shows, the unanimous response was that they watch them for information. Viewers believe the shows portray and they think these shows serve as a channel between population and authorities. S., a 50 years old saleswoman who has been watching Comunidade Alerta since its begi every day because I like to be informed. I reality. I need to know what is going on in the streets. Who is a criminal, a trafficker, he old, who works in maintenance, is the only one that admitted that he also watches the shows f thing communication, June 17 th 2011). the reality they live in, despite the fact that most of them have never actually been victimized. Producers of Comunidade Alerta and Al Amazonas also believe they know what the audience wants and which stories are newsworthy. Elenilza Vieira, director of Al Amazonas, explains people to talk about a certain topic. But the network realized that this format [thematic crime newscasts] would bring more audience. Crime stories ( personal communication, June, 14 th 2011).

PAGE 71

71 Jander Tabosa, from Comunidade Alerta, believes the choice of content goes stories by the impact they c an have on them [the authorities]. We see absurd things, things that we believe authorities and communities should be aware of. We need to alert them, hence the name of the show. What we broadcast is what the population wants to say, what the population w ants to tell the authorities. We show reality, the reality people see on the streets (personal communication, June, 9th, 2011). The preference for graphic images at the cost of greater contextualization is also a major characteristic of the coverage of Co munidade Alerta and Al Amazonas. The extensive use of images of corpses, blood, mutilations, drugs, and weapons is r of Al Amazonas, explains that she also instructs the reporter to go further, to get that extra image that the other they cover. In fact, the videoreporter from Comun idade Alerta, who has been working with Tabosa for seven months, describes his experience as a crime reporter: On my first day i n the field, I had to cover two homicides. It was horrible; was there with me also felt sick and vomited, which is quite normal since the smell [of corpses] is too strong. But this is what we do: we survive on The smell is just too h orrible. (personal communication, June, 8 th 2011) Drug/weapon related crimes have also proven to be newsworthy in Amazonas, representing 19% of the sample (36.5% Comunidade Alerta, 11.9% Al Amazonas). The majority of the homicide stories analyzed mention ed narcotics, either as the result

PAGE 72

72 of drug disputes or addiction, or implying that the murderer was under the influence of drugs. When interviewing criminals, one of the standard questions reporters ask is whether they were addicted to drugs. For example, during while preparing for an interview with a robber, the reporter from Al Amazonas questioned if he used drugs, to the presence of the police authority and there is litt le concern for source confidentiality. After the camera was on, she asked him again, and this time he refused to answer. Her reaction to the refusal was to angrily announce that just minutes ago he had confessed worth mentioning that the story was about an armed robbery at a bus stop, and had no connection with drug trafficking or addiction other than the implied relationship created Conn ecting homicides to drugs also happened in several episodes of Comunidade bank, his name goes to a collection agency. When a citizen owns money to the drug traffickers, h not clear. It seems as if victims and criminals are drug addicts by default, even when the police authority affirms they cannot tell if the crime was a result of drug disputes. Using the PNAD victimization survey, it is impossible to proceed with an interreality comparison and determine if the media portrayal of narcotrafico in Amazonas is exaggerated since PNAD only asks questions about theft, robbery and assault. However, accor ding to Lieutenant Colonel Amadeu da Silva Soares Jnior, a

PAGE 73

73 representative of the Military Police in Amazonas, statistics show that about 65% of the homicides in Manaus are connected to drug activities (Personal communication, June, 29, 2011). Lt. Col. So ares, nevertheless, believes that, more than an attempt to portray real statistics, crime shows depict drug related stories because the audience likes them. make an impa The relationship between authorities and media practitioners also influences the amount of stories on drug seizures in the shows. Tr aditionally, the routine on B razilian newsrooms includes a patrol where the producer checks all police stations and morgues in the city looking for something newsworthy. In Manaus, the situation is no different. The very first procedure in the newsrooms in the morning is to call all police stations in the city and ask if they have anything for the shows. The police authority is the one who defines the limits of the story. A good reporter is the one who builds a close relationship with investigators and poli ce commissioners, which gives them access to the criminals who are arrested. Fabiola Gadelha, reporter from Alo Amazonas, explained that the only constraints to her coverage are the ones set by police authorities. She says: I go as far as I can, always try ing to cross any limit. We have the bulletproof jackets that TV A Crtica provided for when we cover violent conflicts. But overall, the policy authority is the one who establishes the limits of the ationship with the police (personal communication, June 14 2011) M.S., producer of Alo Amazonas, said one of the main difficulties they now face is building relationships with the recently hired police authorities in the state. Without the support of the police and the state secretary of security, it would be very unlikely that

PAGE 74

74 they would have access to the police operations or the criminals inside jails of Amazonas (personal communication, June 13, 2011). The priority given to the police as a source of n ews is one of the main criticisms faced by investigative journalism in Latin America. According to Klahr (2011): Historically, there is a paradox in news coverage of crime and violence throughout Latin America. The media industry and journalists in genera l often rely on official information that is unreliable and unverifiable (and even fictional). We then disseminate this information widely among the population, often treating it more as infota inment and less as serious news . When journalists are atta cked and silenced by the same inefficient and corrupt criminal justice system that they often help legitimize on a daily basis, it highlights the grave paradox that journalists face: they are both facilitators and victims of a repressive system t hat they s hould be challenging (p.18) The stories covered by Al Amazonas and Comunidade Alerta about drug trafficking or homicides have an emphasis on the factual, that is, in the event itself without providing background or follow up. In the case of homicides, th ere are no follow up stories on the investigations. Drug seizures are out of context, and no connection is made between the arrests of those traffickers and the impact to the drug industry in the state, or what are the public security policies for fighting drug trafficking in Amazonas. Rey (2002) identifies as this phenomenon as fruition, regarding to the attractiveness of factual news about violent crimes. Homicides are covered briefly and there is no time to do follow up stories that could have a more rig orous investigation and a less frenetic continuity and leads to the impossibility of reconstructing history in a continuous and coherent way, or identifying the consequenc es of the event covered to the big picture.

PAGE 75

75 In the case of Comunidade Alerta and Al Amazonas, the only contextualization personal or spontaneous, but instead are scripted and, in the case of Al Amazonas, Amazonas explains that he gets the details to construct his comments from the reporters: I call the reporter and, although he/she tries to be impartial, through experience he/she can tell me his/her feeling about the story. There are prove because he has many lawyers and resources; and there are times when the citiz en was condemned but because of small things. The reporter tells me this information so I can form my opinion based on what I already know is the position of the network in those cases. Sometimes I need to make sure with the direction and or with our lawye r [who also participates in th 2011) Fernando Mendes, producer of Comunidade Alerta, is the responsible for the to ments are essential to their contextualization, especially in the case of Comunidade Alerta, where the host serve as an explicit mediator between audiences and authorities. This mediated relationship is part of the agreement between viewers and media prac titioners of the shows. While the reporters are believed to show the reality on the streets, the hosts are the ones that analyze the implications of the story and call for action by authorities.

PAGE 76

76 In conclusion, the results from the content analysis reveal that Comunidade Alerta and Al Amazonas coverage on crime is not representative of official crime statistics. While Al Amazonas has a more diversified coverage, presenting stories on drug trafficking, robbery, assault, and theft, Comunidade Alerta a sho w hosted by politicians covered exclusively homicides and drug related crimes. Nevertheless, homicides are the most newsworthy transgression in Amazonas local TV. In almost half of the victimizations portrayed, the actual victims did not appear in the s tory. When victims appeared, some of their most important characteristics were omitted. Race and occupation were the most omitted information. Among the victims depicted, there was an overrepresentation of darker and younger victims. Although Mapa da Viol ncia shows that young Afro Brazilians are the most victimized group when it comes to homicides, the data from PNAD does reveal an underrepresentation of white adult (30 49 years old) non lethal victimization.

PAGE 77

77 Table 4 1. Number and percentage of stories t hat depicted offense types by program June 2011 Comunidade Alerta Alo Amazonas Combined % n % n % n Homicide** 62.5 10 33.3 14 41.4 24 Motive Money 0 0 21.4 3 12.5 3 Drugs 50 5 7.1 1 4.2 1 Passion 0 0 14.3 2 8.3 2 No reason 0 0 21.4 3 12.5 3 Unknown 50 5 28.6 4 36.7 9 Theft 0 0 11.9 5 8.6 5 Robbery*** 0 0 35.7 15 25.9 15 Drug/weapon trafficking 37.5 6 11.9 5 19 11 Assault 0 0 7.1 3 5.2 3 Total 100 16 100 42 100 58 Percentages based on total number of crime stories in ea ch show. ** Includes attempted murder ***Includes burglary/land invasion Table 4 2. Use of violent images per story by show June, 2011 N Mean Comunidade Alerta 16 2.18 Al Amazonas 42 1.19 Total 58 1.46 p<.00

PAGE 78

78 Table 4 3. Number and p ercentage of TV stories that depicted victims c haracteristics by program June, 2011 Comunidade Alerta Alo Amazonas Combined n=16 n= 42 n=58 Does the vict im appear*? % n % n % n Yes 25 4 47.6 20 41.4 24 No 37.5 6 35.7 15 36.2 21 No victim 37.5 6 16.7 7 22.4 13 Does the victim's body appear**? Yes 77.8 7 33.3 4 52.4 11 No 22.2 2 66.7 8 47.6 10 Characteristics**** Gende r Male 90 9 74.3 26 77.7 35 Female 10 1 14.3 5 13.3 6 Unknown/not depicted 0 0 11.5 4 8.8 4 Age below 18 10 1 11.4 4 11.3 5 18 29 30 3 22.9 8 25 11 30 39 20 2 14.3 5 15.9 7 40 49 10 1 2.9 1 4.5 2 50 59 10 1 2.9 1 4.5 2 60+ 10 1 0 0 2.2 1 Unknown/not depicted 10 1 45.8 15 36.3 16 Race/Ethnicity White 10 1 8.6 3 8.8 4 Black 0 0 2.9 1 2.2 1 Indigenous 0 0 0 0 0 0 Mixed ethnicity/Brown 50 5 34.3 12 37.7 17 Yellow (Asian) 0 0 0 0 0 0 Unknown/not depicted 40 4 54.4 19 51.1 23 Occupation Unemployed 0 0 5.7 2 4.4 2 Builder/Mason 20 2 2.9 1 6.6 3 Commerce 10 1 14.3 5 13.3 6 Police 0 0 8.6 3 6.6 3 Civil servant 0 0 2.9 1 2.2 1 Student 20 2 11.4 4 13.3 6 Ot her 0 0 17.1 6 13.3 6 Unknown/not depicted 50 5 37.2 13 40 18 In the case of homicides, photographs or images of the victim prior victimization ** Lethal cases only **** Excluding stories with no victims, according to PNAD categories

PAGE 79

79 Table 4 4 Number and p ercentage of stories that depicted non lethal victims characteristics* June,2011 Al Amazonas n=23 Does the victim appear? % n Yes 52.2 11 No 47.8 12 Characteristics Gender Male 43.5 10 Fem ale 17.4 2 Unknown/not depicted 39.1 9 Age below 18 4.3 1 18 29 13 3 30 39 0 0 40 49 0 0 50 59 0 0 60+ 0 0 Unknown/not depicted 82.6 19 Race/Ethnicity White 8.7 2 Black 4.3 1 Indigenous 0 0 Mixed ethnic ity/Brown 26.1 6 Yellow (Asian) 0 0 Unknown/not depicted 60.9 14 Occupation Unemployed 0 0 Builder/Mason 0 0 Commerce 17.4 4 Police 8.7 2 Civil servant 4.3 1 Student 20 2 Other 17.4 4 Unknown/not depicted 52.1 12 Com unidade Alerta depicted no non lethal victims

PAGE 80

80 Table 4 5 Portrayal of the victim by r ace N Mean White 4 6 Afro Brazilian 18 4 Not depicted 22 4.27 Total 44 4.31 p<.05 Table 4 6 Portrayal of victimization by gender N Mean Male 34 4.17 Fe male 6 6 p<0.03

PAGE 81

81 CHAPTER 5 CONCLUSION Crime and violence dominate the headlines of daily newspapers and local TV shows in the biggest city of the Amazon region of Brazil. In June of 2011, four gruesome The stories included children murdered by family members, drug victims, decapitations, torture, and domestic violence. But it is on television that local crime coverage acquires its most peculiar characteristics: opinionated pieces, gruesome images, vigila nte hosts, and widespread chaos. More than just audience leaders, these shows reveal much about public perceptions of crime, victimization, and justice, and, indeed, about the culture and politics of Brazil. Overall, crime on TV is by far more violent tha n actual statistics. Programming priority is given to homicides, while robbery and theft, the most common crimes in the state of Amazonas, are largely underrepresented. While these findings might seem obvious, it is the way that crimes and the actors invol ved are portrayed, and the way information is omitted, that can reveal more about the social construction of crime and victimization in the state. Of the 26 states in Brazil, Amazonas is the 17th most dangerous when it comes to homicides. Its rate of 24.8 homicides per 100,000 inhabitants is lower than the national rate of 26.4, and significantly lower than the rate for the northern states of Brazil (32.1). Homicides represent only 0.24% of the crimes in the State, but are the most newsworthy type of victi mization in local TV shows. This finding comes as no surprise to media analysts. Homicides are the most extreme act of violence, and an obviously newsworthy occurrence. The problem with the TV portrayal of homicides in the state of Amazonas is that it is n ot only exaggerated, but also covered out of context. In 36.7% of

PAGE 82

82 the homicides included in the case study of TV crime programs, the motive was never type of explanation that offers an insight into the nature of Brazilian culture. As anthropologist Teresa Caldeira (2000) explains, in the Brazilian society, crime is understood as a affect the way you l ive your life. Caldeira uses that insight to explain the prevalence of lynching, as well as the draconian measures that people endorse as a means to contain what they see as a communicable pathology. In the context of TV crime shows, these culturally base d is accepted, in the minds of many people, as sufficient explanation of a criminal event. The fact that such reasoning would unlikely be advanced as the basis of a crime story in the United States, points to the role that cultural beliefs play in the way in which stories are selected, interpreted, and explained. Corpses and body bags take the place of background information, or any attempt to put victimization into context. The cove rage of homicides thus creates the idea that murder can happen anywhere, to anyone, and for no apparent reason In fact, official statistics show distinct patterns of association with a number of soc io demographic characteristics. Homicides are especially prevalent among younger males and among people of darker skin In contrast to these patterns, homicides on TV are far more random than the profile of actual risks described by PNAD. The idea promoted in the media that homicides are not only a common occu rrence, but can also afflict anyone at any time can generate widespread feelings of insecurity. The media have thus been implicated in fomenting a sense of citizen vulnerability, which, according to some

PAGE 83

83 analysts, leads people in the Latin American cont ext to advocate extra legal solutions, mano dura (get tough) policies, and to vote for populist demagogues who promise easy When it comes to the coverage of non lethal victimizations, Comunidade Alerta and Al Amazonas had different patterns of coverage. Comunidade Alerta covered only drug related stories: drug seizures, gang activities, and arrest of traffickers. Al Amazonas covered robberies, thefts, drug related transgressions, and assault. When comparing the two TV pro grams with PNAD data, Al Amazonas actually provided a relatively representative portrayal of the occurrences of non lethal victimizations. lethal crime, is also the most common non lethal crime on the streets. Comun idade Alerta, on the other hand, focused its coverage on homicides and drugs, using twice as many violent images per story than Al Amazonas. The creators of Comunidade Alerta who are also hosts and politicians are on a personal televised crusade agains t drugs in the state. They spend most of their airtime explaining how the drug industry has grown in Manaus and promoting mano dura policies to apprehend drug traffickers. During the weeks I accompanied the show, the hosts openly supported death penalty, r ejected the idea of civil rights for criminals, and advocated arming citizens. They repeatedly claimed that the state disarmed citizens yet the jail system. The viewers of the programs often repeated these discourses when I questioned them as to why they thought Manaus was a dangerous city. No causal connection can be inferred from my interviews if only because it might well be that

PAGE 84

84 citizens that already hold such views and support the policies advocated by TV hosts are people who prefer to watch shows that share their views on crime. Nonetheless, the shared discourse between the program hosts (and their producers) and viewers of crime shows in the state is an observation consistent with the idea that the media influence public perceptions. The creators of Comunidade Alerta openly state that they are performing a public service to the city. They do not see a problem with being the mediators between the audiences. Compared to televised crime shows in the United States, those in Brazil seem to be expressly embedded in the logistics and the politics of crime control. Al Amazonas also frequentl y promotes mano dura policies, especially when it comes to granting rights to criminals. All the journalists interviewed showed no concern about the image rights of the criminals. The host of Al Amazonas explained that, when human rights groups complain a bout the way journalists cover criminals in the state, the population usually protests against such allegations (personal communication, June 10th, 2011). The police have a vested interest in displaying the weapons and drugs that have been seized, and sho wing the criminals that have been arrested, thereby demonstrating the efficacy of their operations in their fight against trafficking in the state. The easy access that journalists have to jails, and the heavy reliance on police authorities as sources of i nformation, mean that stories give greater emphasis to criminals, their weapons and their connections to the drug trafficking, as well as gruesome depictions of

PAGE 85

85 the deceased compared to the emphasis on the characteristics of the victims, the causes of the event, or the context in which it occurred. These characteristics are not particular to local TV in Amazonas, but are common to all crime coverage in Latin America. According to Rey (2002): What viewers or readers will remember are few generic schemes of crime: offenders exposed to the cameras trying to hide their identities, places with traces of the crime, anonymous bodies, seized weapons displayed to the public, prisons crowded with criminals (Rey, 2002, p.25) The stories emphasize a single moment in t he course of events that led to the crime and there is no background reference that c an provide the viewer with some understanding of the bigger picture. In the case of the superficiality of the coverage on drugs one of the main crimes depicted by Comuni dade Alerta and Al Amazonas Sierra (2011) notes: What the public ends up learning about the drug trade is but the tip of the iceberg of a phenomenon whose complexities barely become mere brushstrokes in the vast mural of information. The topic of ille gal drugs is reduced to the drug traffickers and dealers, essentially focusing on police efforts and the myths propagated by state offices. The topic of violence is concentrated in a handful of clichs repeated once and again until they become shared knowl edge (p.38) inform the audience of what is necessary to avoid being victimized. Crime on TV in Amazonas is not only exaggerated, but its victims are also random and ano nymous Victims of the crime were portrayed in only 41.4% of the total stories. Even when victims are portrayed, it is often impossible to determine their race or gender, due to the conditions of their body. When victims do appear on the screen, r elevant i nformation such as occupation and age of the victim, are not reported in almost 40% of the cases.

PAGE 86

86 When the characteristics of victims are revealed, my case study shows that homicides on TV are mainly young Afro Brazilian (mostly brown) males, a finding t hat is in accordance to the data from Mapa da Violncia. However, when it comes to non lethal victimizations, there is an underrepresentation of black and white victimization, precisely the groups at highest risk of being victimized by non lethal victimiza tions on the streets. Adults (30 49 years old) and females are also underrepresented as the victims of robbery, theft, and assault. The lack of female victimization on TV is a phenomenon that has been much studied in feminist criminology. According to Mel oy and Miller (2009): The masculine perspective that we are describing tends to underreport male on female violence while highlighting the atypical event: extreme acts of violence by or against women. Most crimes against women are considered ordinary and c ommon from a masculine perspective. They rarely become news items. Up to 95% of all reported crimes against women receive no media attention. (p.31) In Comunidade Alerta and Al Amazonas, female victims were actually portrayed more positively than males, although there were only six cases involving females (four homicides and two assaults). As PNAD and Mapa da Violncia show, these cases are not representative of official female victimization statistics. Males are twice as likely as females to be a victi m of assault, and 92% of the homicides reported in Mapa da Violncia had male victims. However, the six females portrayed in Comunidade Alerta and Al Amazonas were victims of these predatory crimes, rather than theft or robbery, the two most common crimes involving female victims, according to PNAD. The media depicts a reality in which women are more likely to become victimized by homicides or assaults, crimes that predominantly have male

PAGE 87

87 victims, when, in fact, females are at greater risk of becoming vict ims of theft and robbery. With respect to race, the underrepresentation of white victims is in contrast to most similar studies conducted in the US media. In the United States, media victims are usually whiter, younger, and more affluent compared to the a ctual statistics (Dixon & Linz, 2000; Jewkes, 2004). US portrayals of race and social groups in crime stories are The ethnic blame discourse frames problem behavior committed by e thnic Others (e.g., Blacks and Latinos) as intergroup conflict and accentuates the harmful effects of the behavior for the in group (e.g., whites). As a result, Black and Latino perpetration of crime, particularly if Whites are victims, may occur frequentl y on news programs. Whites, however, should occupy benign or helpful roles on television news programming (p.132) unwillingness to be portrayed in Comunidade Alerta and Al Amaz onas challenges the assumption that being a victim is understood as a benign role in the crime phenomenon of violent crimes. Even when there is no evidence that the specific story shown has any relation to narcotics, the hosts often make comments about how drugs are corrupting the youth. random than it is in reality. Anyone could be a victim of crime on TV, especially when the only part of the victim portrayed is a body bag or a pool of blood. Victims of robbery, assault, and theft also refuse to appear, giving space to the excessive portr ayal of the criminals, the police, and their methods.

PAGE 88

88 The emphasis that local TV coverage gives to criminals and to police actually suggest that viewers, rather than watching crime shows to be informed, actually watch them to reinforce their already cons tructed ideas about crime. Additional information that could be provided e.g., about the victims, the context, and the causes of the event This research is mainly descriptive of two speci fic TV shows. Future studies using a representative sample of several shows across the country are necessary to draw generalizable conclusions about crime coverage in Brazil. Nevertheless, this study reveals several patterns of crime coverage by Comunidade Alerta and Al Amazonas that are similar to other studies of crime coverage in Latin America (Klahr, 2011; Rey, 2002; Sierra, 2011), which find that crime news is superficia l, that programs rely heavily on authorities for information, that the stories are depicted in a highly dramatized fashion, often explaining events in terms of the simple dichotomy between good and evil. The effect of crime shows on the people in Brazilian society is yet to be studied, although the discussion between media portrayals of been the focus of several studies in the United States. While this relationship remains unexplored in Brazil, the Ministry of Justice in establishes a rating system for television shows, presumably to protect vulnerable audiences from content that could damage Comunidade Alerta are broadcasted during the lunch hour.

PAGE 89

89 Citi zens get this information school. The constant political participation and election to offic e hosts of crime shows, in particular Comunidade Alerta, insert themselves into the story, identify criminals, motives, and advocate certain pol icies and punishments. While the journalism, these media figures have gone well beyond reporting and have advanced into the field of politics. There is little pretense that t he facts are revealed in an obj ective manner, and, in any case audiences do not seem to value objectivity. Instead people value the ability of these shows to solve their problems and their effectiveness in calling attention to the problems people face in their communities Feeling trapped by crime and cor ruption, citizens turn to media figures to solve violence, even after the 2004 scandals that could have put an end to the success of crime shows in the state. watches both shows daily (personal communication, June 17th, 2011). Even the names of the shows Comunidade Alerta (Community Alerts) and Al Amazonas (Hello, Amazo nas) refer to their role as intermediaries between pe action. This space where TV hosts mediate between the commun ity and offices of the state can be understood as a manifestation of the lack of democratic representation in Brazil. In this sense, TV crime shows step in to repl ace the weaknesses in the institutionalized means of interest representation. If the character and content of crime

PAGE 90

90 shows reflect Brazilian culture, it is also evident that Brazilian politics influence the role played by televised crime programs. It is fr equently argued that the exaggerated attention given to gruesome images of crime can cause audiences to become desensitized to the events depicted. and news coverag e of the war, event driven as all news is, has become the point when a certain amount of exposur e to violent images, audiences d o not bec ome more fearful but, inste ad, become insensitive to issues regarding crime and are no longer shocked by violence. We are thus faced with two seemingly contradictory hypotheses. One, in accordance with much literature on crime portrayal in the United States, predicts that the dramat ized and distorted images of criminality in the media foster an unrealistic fear of crime; the other, in accordance with Guillermoprieto, contends that the televised depiction of criminality desensitizes people to the crime and violence around them. Which phenomenon predominates is likely to vary by context, and probably by the kind of crimes involved. Perhaps both processes are at work, as people become increasingly sensitized to some features of criminality yet also come to acc ept other features as routi ne. Whatever the case, these observations point to important new directions for future research.

PAGE 91

91 APPENDIX A CODING GUIDE: PORTUGUESE A unidade a ser analizada ser cada reportagem individualmente. As reportagens devem abordar temas pelo menos uma trang resso (crime). Denncias sobre outros problemas (falta de saneamento, poltica, etc) no devem ser codificados. O codificador deve analisar a reportagem (externas), bem como os comentrios do apresentador SOMENTE sobre o tema da reportagem. O total do tem po de cada histria inclui a externa do reprter e os comentrios do apresentador sobre o tema. Comentrios soltos sem relao com as reportagens no devem ser codificados. Todos os campos devem ser preenchidos. Instrues: codifique todas as histrias que atendem a seguinte especificao: 1) relacionadas a crime(incluindo todos os tipos de transgresso). Todas as outras histrias devem ser codificadas 99. 1. Caso ______ (comear com 101 para Comunidade Alerta e 501 para Alo Amazonas. Exemplo: 1011, 5012) 2. Nome do programa: (1) Comunidade Alerta (2) Al Amazonas 3. Data: tudo em nmeros (DDMMAA)

PAGE 92

92 4. O comprimento total da histria (em minutos, indicar quando comea e quando acaba, incluindo comentrios do apresentador sobre o caso)________ 5. Qual o principal tipo de infrao representada pela histria? (Caso o infrator tenha mais de uma infrao, marque a que o programa deu maior enfoque) 1. Assassinato 2. Fraude / extorso 3. Fuga 4. Furto (sem agresso fsica) 5. Roubo (furto com emprego de violncia) (inclui se : assalto mo armada) 6. Porte/trfico de armas ilegais 7. Terrorismo 8. Desaparecimento 9. Morte inexplicada 10. Estupro 11. Sequestro 12. Formao de quadrilha 13. Abuso sexual infantil 14. Tentativa de homicdio 15. Trfico de drogas 16. Apreenso de drogas 17. Invaso de terras 18. Agresso f sica.

PAGE 93

93 19. Outros. Especifique ____________ *No caso de reportagens sobre invaso de reas/acampamento/desocupao de terras, o infrator o invasor, exceto nos casos de violnc ia policial contra os invasores 6. Qual a arma do crime? 1. Faca 2. Teado 3. Arma de fo go 4. Other___________________________ 5. No retratada 6. No se aplica 7. Qual foi o motivo do crime? 1. Dinheiro 2. Dependncia de drogas (para comprar drogas; vtima de dvidas por drogas) 3. Disputas de trfico (entre traficantes/gangs) 4. Sexo 5. Passional 6. Religio 7. Efei to de entorpecentes (sem razo outra alm do efeito de narcoentorpecentes)

PAGE 94

94 8. 9. Outros ___________________________ 10. Desconhecido / no descrita 8. Em que zona ocorreu o crime? 1. Urbano 0. Rural 9. Em caso de crime urbano (Se 8 = 1) Em que zona ocorreu o crime (ver tabela dos bairros em anexo)? 1. Norte 2. Sul 3. Leste 4. Oeste 5. Centro oeste 6. Centro sul 7. Desconhecida 10. Qual o local do crime? 1. Residncia da vtima 2. Residncia do infrator 3. Estabel ecimento commercial

PAGE 95

95 4. Via pblica 5. Estabelecimento de ensino 6. Transporte coletivo 7. Ginsio ou estdio esportivo 8. Outros 9. No descrito 11. O autor do crime aparece na histria (incluindo foto)? 1. Sim 2. No Quantos infratores so retratados (colocar 99 quando desconhe cido) _____ Sobre as caractersticas do infrator: preencher um para cada infrator 12. Atual situao: 1. Preso 2. Desaparecido 3. Morto 4. Outro/Em liberdade provisria 5. No se aplica 13. Sexo: 1. Masculino 2. Feminino

PAGE 96

96 3. Desconhecido/No citado 14. Idade 1. Inferior a 18 2. 18 29 3. 30 39 4. 40 49 5. 50 59 6. 60+ 7. Desconhecido / No representado 15. Etnia 1. Branco 2. Negro 3. Indgena 4. Etnia mista/Pardo 5. Amarelo 6. Desconhecido / No representado 16. Profisso 1. Desempregado 2. Construtor/Pedreiro 3. Agricultor

PAGE 97

97 4. Comerciante 5. Polcia 6. Foras Armadas 7. Servios d omsticos 8. Servidor pblico 9. Servios informais (ambulante, bico, etc) 10. Estudante 11. Outros (especificar) _______________________________ 12. Desconhecido / No representado 17. Como o autor do crime aparece na histria? 1. Preso com rosto mostra 2. Solto (imagens distncia) 3. Preso cobrindo o rosto 4. Apenas foto 5. No aparece 6. No se aplica Sobre a(s) vtima(s) 18. A vtima viva ou antes de morrer (em casos de homicdio) aparece na histria? (excluindo cadaver, saco de cadaver, caixo, pedaos do corpo, etc. Incluindo foto da vtima e vtima viva)

PAGE 98

98 1. Sim 2. No 3. No se aplica (sem vtima) 19. A familia da vitima aparece na histria? 1. Sim 0. No 19. O corpo (cadaver/partes do corpo, excluindo caixao fechado/saco de cadaver fechado) da vtima aparece na histria? 1. Sim 2. No 3. N o se aplica 20. Se sim, como o corpo da vtima aparece? 1. Muito visivel 2. Apenas um detalhe 3. Embaado 4. Outra forma ________________ 5. No se aplica Como a vtima descrita? (Preencher um para cada vtima) 21. Sexo:

PAGE 99

99 1. Masculino 2. Feminino 3. Desconhecido/ No descrito 22. Idade 1. menos de 18 2. 18 29 3. 30 39 4. 40 49 5. 50 59 6. 60+ 7. Desconhecido / no descrita 23. Etnia 1. Branco 2. Negro 3. Indgena 4. Etnia mista/Pardo 5. Amarelo 6. Desconhecido / no descrita 24. Profisso 1. Desempregado

PAGE 100

100 2. Construtor/Pedreiro 3. Agricultor 4. Comrciante 5. Polcia 6. Foras Arm adas 7. Servios domsticos 8. Servidor pblico 9. Servios informais (ambulante, bico, etc) 10. Estudante 11. Outros (especificar) _______________________________ 12. Desconhecido / No representado 25. Por favor, especifique outros adjetivos atribudos vtima Sobre a polcia 26. Como a polcia aparece na histria? 1. No aparece 2. Somente citado pelo apresentador / reprter 3. Imagens da polcia, mas no fala 4. Imagens e depoimento da polcia (incluindo investigadores e delegados falando como fontes) 5. Outra forma________________ __

PAGE 101

101 6. No se aplica 27. H qualquer meno polticas pblicas de segurana na histria? (incluindo secretaria/secretrio de Segurana, prefeito, governador, etc)? 1. Sim 2. No Sobre as imagens usadas 28. As imagens mostram? Table A 1. Images Portrayed Portugu ese SIM (1) NAO (0) Arma branca Arma de fogo Ferimentos Saco de cadaver/Caixo Partes de corpo Sangue Drogas Munio Outros (especificar) Em uma escala de 1 a 7 1 Negativo

PAGE 102

102 4 Neutro ( marcar neutro quando no for citado e xplicitamente) 7 Positivo 99 No se aplica (marcar quanto no for relacionado com a reportagem) 29. Qual a imagem geral da vtima na histria?__________ 30. Qual a imagem geral da polcia na histria?___________ 31. Em relao a segurana, como est a situao geral da cidade descrita?___________

PAGE 103

103 APPENDIX B CODING GUIDE : ENGLISH The unit of analysis is each individual story The story must address at least one transgression (crime). Stories about other issues (lack of sanitation, politics, etc .) should not be coded. The coder shall code the story and the comme nts of the host ONLY on that specific case The total length of each story includes the story and host s comments on the subject. Comments unrelated to the stories should not be coded. All fields must be filled. Instructions: code all the stories that meet the following specification: 1) related to crime (including all t ypes of transgression stories ). All other stories should be not be coded 99. 1. Case ______ (start with 101 for Com unidade Alert a and 501 for Alo Amazon as) Example: 1011, 5012) 2. Name of program: (1) Comunidade Alerta (2) Al Amazonas 3. Date: all in the numbers (DDMMYY) 4.________ The total length of history (in minutes indicate when it starts and ends, including the comments on the case ) 5. What is the main type of offense represented by the story? 1. Murder 2. Fraud / extortion 3. Escape

PAGE 104

104 4. Theft (without physical aggression) 5. Robbery ( with use of violence) (includes the following: armed robbery) 6. Possession / tra fficking of illegal arms 7. Terrorism 8. Disappearance 9. Unexplained death 10. Rape 11. Kidnapping 12. Conspiracy 13. Child sexual abuse 14. Attempted murder 15. Drug traffic 16. Drug bust 17. Invasion of land /property 18. Assault 19. Other. Specify ____________ In case of invasion on areas / camp / clearing of land, the offender is the INVADER except in cases of polic e violence against the invaders. 6. What is the weapon of crime ? 1. Knife 2. Machete 3. Firearm

PAGE 105

105 4. Other___________________________ 5. It is not portrayed 6. Not applicable 7. What was the motive? 1. Money 2. Drug addiction (to buy drugs, victim of drug debts) 3. Drug disputes (between dealers / gangs) 4. Sex 5. Passional 6. Religion 7. E ffect of drugs (for no reason other than the effect of narcoentorpecentes) 8. For no apparent reason (including "pure evil", "wanting only to do evil", etc.) 9. Other ___________________________ 10. Unknown / not described 8. Area in which the crime occurred? 1. Urban 0. Rural 9. In case of urban crime (If 8 = 1) In what zone the crime? 1. North 2. South

PAGE 106

106 3. East 4. West 5. Midwest 6. South Central 7. Unknown 10. Where did the crime happen? 1. Residence of the victim 2. Residence of the offender 3. Commercial establishment 4. Public 5. Educational establishment 6. Public transportation vehicle 7. Gymnasium or sports arena 8. Other 9. Not described 11. Do the criminals appear in history (including photo)? 1. Yes 2. No How many offenders are portrayed ? (insert 99 when unknown) _____ About the characteristics of the offender: fill one for each offender

PAGE 107

107 12. Current situation: 1. Arrested 2. Disappeared 3. D ead 4. Other _____________ 13. Gender: 1. Male 2. Female 3. Unknown / not mentioned 14. Age 1. Less than 18 2. 18 29 3. 30 39 4. 40 49 5. 50 59 6. 60 + 7. Unknown / not represented 15. Race (according to PNAD categories) 1. White 2. Black

PAGE 108

108 3. Indian 4. Mixed Ethnicity / Brown 5. Yellow 6. Unknown / not represented 16. Profession (according to PNAD categories) 1. Unemployed 2. Builder / Mason 3. Farmer 4. Commerce 5. Police 6. Armed Forces 7. Domestic Services 8. Public servant 9. Informal services 10. Student 11. Other (specify) _______________________________ 12. Un known / not represented 17. How does the offender appear in the story? 1. Arrested, showing the face 2. Not arrested/Image from distance 3. Arrested covering face 4. Picture only

PAGE 109

109 5. Not appear 6. Not applicable Please describe other adjectives used to describe the of fender: About (s) victim (s) 18. Does t he victim alive or prior to victimization (in cases of murder) appear in the story? (Excluding corpse bag corpse, coffin etc. Including picture s and videos of the victim ) 1. Yes 2. No 3. Not applicable (no victim) 19. Does the victim's family appear in history? 1. Yes 0. No 19. Does t he body (cadaver / body parts, excluding coffin/ sealed body bag) appear in the history of the victim? 1. Yes 2. No 3. Not applicable 20. If so, how does the victim's body appear ?

PAGE 110

110 1. Very vis ible 2. Only detail s (some parts) 3. Blurred 4. Other ________________ 5. Not applicable (Complete one for each victim) 21. Gender: 1. Male 2. Female 3. Unknown / not described 22. Age 1. less than 18 2. 18 29 3. 30 39 4. 40 49 5. 50 59 6. 60 + 7. Unknown / not described 8. 23. Color (according to PNAD categories) 1. White

PAGE 111

111 2. Black 3. Indian 4. Mixed Ethnicity / Brown 5. Yellow 6. Unknown / not described 24. Profession (according to PNAD categories) 1. Unemployed 2. Builder / Mason 3. Farmer 4. Commerce 5. Police 6. Armed Forces 7. Domestic Services 8. Pub lic servant 9. Informal services 10. Student 11. Other (specify) _______________________________ 12. Unknown / not represented 25. Please specify other adjectives attributed to the victim About the police

PAGE 112

112 26. How does the police show in the story? 1. Not appear 2. On ly mentioned by the presenter / reporter 3. Images of the police, but does not speak 4. Images and testimony of the police (including speaking as sources) 5. Other 6. Not applicable About the images 28. The imag es show? Table A 2 Images portrayed English Y ES (1) No (0) Melee Weapon Firearm Injuries Body bag / Coffin Body parts Blood Drugs Ammunition Other (specify)

PAGE 113

113 On a scale from 1 to 7 1 Negative 4 Neutral ( mark when not explicitly mentioned) 7 Positive 9 9 Not applicable ( mark when not related to the story ) 29. What is the overall ima ge of the victim in the history ? __________ 30. W hat is the overall ima ge of the police in the history ? ___________ 31. Regarding safety, how is the general situation of the city described ? ______

PAGE 114

114 APPENDIX B PICTURES Figure B 1. Al Amazonas crew interviewing criminal in police station in Manaus 06/14/2011 Photo courtesy of Rachel Reis Mourao.

PAGE 115

115 Figure B 2. Reporter from A l Amazonas meeting with police. Photo cou rtesy of Rachel Reis Mourao. Figure B 3. Al Amazonas equipment mixed with crime evidence Photo courtesy of Rachel Reis Mourao.

PAGE 116

116 Figure B 4 Host of Comunidade Alerta discussing a homicide with Raimundinho, the Photo courtesy of Rache l Reis Mourao.

PAGE 117

117 LIST OF REFERENCES Albuquerque, A. (2000). Um outro quarto poder: Imprensa e compromisso poltico no brasil. Contracampo, 4 125 145. Associated Press. (2009). Murder for ratings? cops suspect TV host. Retrieved 10/09, 2011, from http://www.cbsnews.com/stories/2009/08/11/world/main5235384.shtml Barak,G.(1994) Media, process, and the social construction of crime New York: Garland. Blankson, I. A ., & Murphy, P. D. (2008). Negotiating democracy: Media transformations in emerging democracies. New York NY : State Univ ersity of New York Press Bucci, E. (1996) Brazil em Tempo de TV. So Paulo Brazil : Bomtempo Editora Caldeira, T. P. (2000). City of W a lls: Crime, segregation, and citizenship in So Paulo Los Angeles, CA: Univ ersity of California Press Carrabine, E. (2008). Crime, culture and the media London, UK: Polity Press Car valho, G. (2008). A mdia e o crime: Que relao esta? Universidade T oledo de Presidente Prudente). Intertemas, 15 (15) (Online) Retrieved from http://intertemas.unitoledo.br/revista/index.php/Juridica/article/view/601/6 16 Cavender, G., & Bond Maupin, L. (1993). Fear and loathing on reality television: An Sociological Inquiry, 63 (3), 305 317. Ceobanu, A. M., Wood, C. H., & Ribeiro, L. (2011). Crime victimiz ation and public support for democracy: Evidence from Latin A merica. International Journal of Public Opinion Research, 23 (1), 56. Correa, N. (2004, October 10). Xerifes eletronicos. A Critica, pp. A3. Costa, A. C. (2008). Mdia e poder: A organizao po ltica da TV rio negro. Unpublished Bachelor of Arts Social Communications U niversidade Federal do Amazonas. Dammert, L., & Malone, M. F. T. (2003). Fear of crime or fear of life? P ublic insecurities in Chile. Bulletin of Latin American Research, 22 (1), 79 101. Dixon, T. L., & Linz, D. (2000). Overrepresentat ion and underrepresentation of African Americans and L atinos as lawbreakers on television news. Journal of Communication, 50 (2), 131 154.

PAGE 118

118 Dixon, T. L., & Linz, D. (2000). Race and the misrepresentat ion of victimization on local television news. Communication Research, 27 (5), 547. Dowler, K. (2003). Media consumption and public attitudes toward crime and justice: The relationship between fear of crime, punitive attitudes, and perceived police effectiv eness. Journal of Criminal Justice and Popular Culture, 10 (2), 109 126. Garzn, G. V. (2008). Mafia & co: The criminal networks in Mexico, Brazil, and C olombia Woodrow Wilson International Center for Scholars. Retrieved from: http://www.wilsoncenter.org/sites/default/files/Mafia%20and%20Company.pdf Greer, C. (2010). Crime and media: A reader. New York, NY: Routledge. Guillermoprieto, A. (2010, 07/30/2010). The murde rers of Mexico. The New York Review of Books Retrieved from http://www.nybooks.com/articles/archives/2010/oct/28/murderers mexico/?pagination=false Gus mao, D. (2004, June 6). Os campeoes de audiencia. O Estado do Amazonas, pp. P1. Hale, C. (1995). Fear of crime: A review of the literature. International Review of Victimology, 4 (2), 79 150. Hall, S., Critcher, C., Jefferson, T., Clarke, J. and Roberts, B. (1978) Policing the Crisis. London, UK: Macmillan Hamilton, F. A. (2009). Encenaes sobre a criminalidade. Revista ECO POS, 12 (2) Henn, R., Oliveira, C., Wolff, M. P., & Conte, M. (2011). Crime in media: An interdisciplinary research. Brazilian Journa lism Research, 1 (2) 159 176. Jewkes, Y. (2004). Media and crime (Key Approaches to Criminology). London, UK: Sage Publications Ltd. Katz, J. (1987). What makes crime news. Media, Culture and Society, 9 (1), 47 75. Kehl, M.R (2004) One people, one head, o ne nation. Retrieved from: http://www.mariaritakehl.psc.br/resultado.php?id=45 Klahr, M. L. (2011). Information that makes issues invisible In: Coverage of Drug Trafficking and Organized Crime in Latin America and the Carribbean report. Austin, TX: Retrieved from http://knightcenter.utexas.edu/ebook/cove rage drug trafficking and organized crime latin america and caribbean LaMay, C. L., & Dennis, E. E. (Eds.).(1995). The culture of crime New Jersey, NJ: Transaction.

PAGE 119

119 Mattos, S. (2002) A Historia da Televiso Brasileira. Petropolis, R J, Brazil: Editora V ozes. M azotte, N. (2011, 11/08/2011). Brazilian cameraman caught in the crossfire while cov ering police operation outside Rio de J aneiro Knight Center for Journalism in the Americas Blog Retrieved from http://knightcenter.utexas.edu/blog/brazilian cameraman caught crossfire while covering police operation outside rio de janeiro Mello Jorge, M., & Comi sso Nacional de Populao e Desenvolvimento. (1998). Como morrem nossos jovens. Jovens Acontecendo Na Trilha Das Polticas Pblicas, 209 292. Meloy, M. L., & Miller, S. L. (2009). Words that wound: Print media's presentation of gendered violence. In D. H umphries (Ed.), Women, violence, and the media: Readings in feminist criminology (29 56) Lebanon, NH: Northeastern Univ Pr. Ministrio da Justia. (2007 July 26 ). Nova portaria regulamenta classificao indicativa. Portal Ministerio da Justica. Retrieve d from http://portal.mj.gov.br/data/Pages/MJB1F4AB69ITEMIDF8C97B73DC1E40EABB3 C5A3C877E4843PTBRNN.htm Morgan, M., Shanahan, J., & Sig norielli, N. (2009). Cultivation processes. Media Effects: Advances in Theory and Research, New York, NY: Routledge. Muraskin, R., & Domash, S. F. (2007). Crime and the media. headlines vs. reality. New Jersey, NJ: Upper Saddle River Oliveira, D. D. Jorn alismo policial, gnero e modo de endereamento na televiso brasileira. Colquio Internacional Televisao e Realidade. Available at: http://www.tverealidade.facom.ufba.br/coloquio%20textos/Dannilo%20Duarte.pdf Prillaman, W. C. (2003). Crime, democracy, and development in Latin America. Policy Papers on the Americas, 14 (6), 1 30. Rey, G. (2005). El cuerpo del delito. Bogot, Colombia: Centro De Competencia En Comunicacin Para America Latina. Ribeiro, L., & Wood, C. H. Crime, fear, and violence in Latin A m erica: Issues, data, and definitions Manuscript submitted for publication. Riffe, D., Lacy, S., & Fico, F. G. (2005). Analyzing media messages: Using quantitative content analysis in research. Mawah, NJ: Lawrence Erlbaum. Rockwell, R. (2007). Vestiges of authoritarianism. Negotiating Democracy: Media Tranformations in Emerging Democracies. Albany, NY: State University of New York Press.

PAGE 120

120 Santos, S.; Capparelli, S. (2005) Coronelismo, Radiodifusao e voto: A nova face de um velho conceito. In Brittos, V.C; Bo lano, C.R. (Eds) Rede Globo: 40 anos de poder e hegemonia Sao Paulo Brazil : Paulus Editora Shanahan, J., & Morgan, M. (1999). Television and its viewers: Cultivation theory and research New York, NY: Cambridge Univ ersity Press Sierra, A. (2011). Is ther e an independent journalistic narrative on drugs? The strange paradoxes of drug coverage in the news. In: Coverage of Drug Trafficking and Organized Crime in Latin America and t he Carribbean report Retrieved from http://knightcenter.utexas.edu/ebook/coverage drug trafficking and organized crime latin america and caribbean Simoes, C.F; Mattos, F. (2005) Elementos histor ico regulatorios da televisao brasileira. In Brittos, V.C; Bolano, C.R. (Eds) Rede Globo: 40 anos de poder e hegemonia Sao Paulo Brazil : Paulus Editora Sodr, M. (1984). O monoplio da fala: Funo e linguagem da televiso no Brasil. So Paulo, Brazil : V ozes. Souza, E. R., Njaine, K., & Minayo, M. C. S. (1996). Qualidade da informao sobre violncia: Um caminho para a construo da cidadania. Cadernos do Programa De Ps Graduao Em Cincia Da Informao, 2 104 112. Truman, J. L. (2011). Criminal vict imization Washington, DC: Diane Publishing. Waiselfisz, J. J. (2011). Mapa da violncia 2011. Os Jovens do Brasil. Retrieved from: http://www.sangari.com/mapadaviolencia/mapa2011.html Za firau, S. (2009). Audience knowledge and the everyday lives of cultural producers in hollywood. In V. Mayer, M. Banks & J. T. Cadwell (Eds.), Production studies: Cultural studies of media industries ( ebook location 4597) New York, NY: Taylor & Francis.

PAGE 121

121 BIOGRAPHICAL SKETCH Rachel Reis Mourao was born in Manaus, Amazonas, Brazil where she obtained ournalism from Universidade Federal do Amazonas. Mourao worked for a year as a journalist in Ma naus before starting the Master in Latin American Studies program at the University of Florida. She will start her PhD program in journalism a nd wants to pursue a career in a cademia.