Journal of Criminal Justice and Popular Culture

Journal of Criminal Justice and Popular Culture
, 9(1) (2001) 1-14



Wes Shipley
Gray Cavender
Arizona State University


Politicians and the public alike blame an array of social problems on movies, which they claim are increasingly violent and graphically violent. In this paper, we analyze the five top-grossing films in a one year period in each of the past four decades (i.e., 1964, 1974, 1984, and 1994). Our analysis, which is based on a careful coding of these films, focuses on acts of violence, graphic violence, and death. We find that violence, graphic violence, and death increases across the four decades, although not in a perfectly linear trend. We also find that the violence is limited to particular genres.


     Since the beginning of the film industry in the early 1900s, Americans have had a love/hate relationship with the movies. Movies are revered as one of our most popular media forms. Many newspapers provide a running account of the gross receipts of the week's top films, and, for many fans, movie stars are the American equivalent of royalty. At the same time, however, movies are reviled as dangerous, especially to youth. They were a prime suspect in the finger-pointing that followed the Columbine tragedy. More recently, the U.S. Federal Trade Commission blasted the film industry for marketing R-rated movies to minors.

     For critics of the film industry, there are two major issues. First, critics argue that movies may instill antisocial values in the viewing audience, especially a youthful audience. They suggest that movies and other popular media convey messages that affect our behavior. The concern is that who we are is shaped in a negative way by these messages.

     The second criticism and the focus of our analysis is the concern that movies are too violent. This criticism includes the twin claims that movies are increasingly violent and that the graphic nature of that violence is increasing as well.

     Obviously, these criticisms represent two sides of the same coin: what we see at the movies shapes our behavior, and what we see is more and more violence. For example, kids who see too many violent, criminal images at the movies may commit violent crimes, either tomorrow or in the future. Specifically, critics contend that violent crime has risen over the past several decades in the U.S. They blame the increase in crime and violence on a number of criminogenic (i.e., crime generating) factors, but chief among them are popular media such as the movies. It is no coincidence, critics argue, that the increasing level of violence in society parallels the increasingly violent content of films.

     Both criticisms are so commonplace as to verge on being assumptions. In reality, of course, these criticisms raise empirical questions. In this paper, we address the second criticism in a rather straightforward manner. We ask, "Are movies really becoming more violent?" "Is the graphic nature of film violence on the increase?" To answer these questions, we analyze the violent content of the five top-grossing films in one year intervals across four decades (i.e., 1964, 1974, 1984, and 1994).

     In the sections that follow, we first discuss the putative relationship between film and violence (End Page 1). We also discuss the claim that movie violence is on the increase. Next, we explain our method (i.e., how we chose the films and operationalized violence and graphic violence). We then present our findings about the violent content of films across these four decades. Finally, we discuss the implications of our findings.


     We have long been a nation that is obsessed with pop culture. Commentators suggest that the media that comprise pop culture help to shape our identity, both individually and as a nation (Anderson 1990; Bailey and Hale 1998). This is variously seen as both a good and a bad thing.

     Movies are one of the most popular and most enduring of our media. Once they began in the early 1900s, movies enjoyed an explosive popularity. In 1904 there were no movie theaters in the U.S. By 1910, there were 5000 theaters, and the moviegoing audience numbered 26 million patrons per week, about one-fifth of the population (Gomery 1992:19-20). Americans clearly had discovered a medium that was dear to their hearts.

     Indeed, movies even help to define what it means to be an American. In the early 1900s, for example, recent immigrants were among the most loyal moviegoers, and the medium was a key component of their socialization (Gomery 1992:21). Similarly, women became loyal fans in the 1910s, and movies helped to define the "New American Woman" in U.S. society (Gomery 1992:31). Beginning in the late 1950s and thereafter, movies became a defining purveyor of youth culture (Snyder 1995).

     However, the very fact that the movies helped to shape the identities of these populations prompts some critics to condemn them. The film industry has been blamed for antisocial behaviors that range from sexual promiscuity to un-American activities to juvenile delinquency (Krutnik 1991; Kidd-Hewitt 1995; Clarens 1997; Todd 2000). Of course, movies are not unique in experiencing such criticism. Over the years, many popular media have been blamed for antisocial behaviors. Critics have posited relationships between comic books and juvenile delinquency (see Nyberg 1998), between music and drugs (see Gray 1995), and between television and violence (Gerbner and Gross 1976; see Surette 1998).

     Despite these broad condemnations of popular media, movies and, more recently, television are frequent targets of criticism. And, although the criticisms cover a variety of behaviors (e.g., sexual and other social mores), complaints about crime and violence perennially underlie the attacks on movies and on television as well (Allen, Livingstone, and Reiner 1997; Beckett and Sasson 2000; Todd 2000; Rafter 2000).

     While criticism of crime and violence in film is not new as early as 1917 movies were blamed for juvenile delinquency (see Allen, Livingstone, and Reiner 1998:54 ) the amount and the volume of criticism has been pronounced over the past four decades. Given these deeply felt concerns, not surprisingly there literally are thousands of scholarly studies of the relationship between popular media and violence (Livingstone 1996; Taylor and Willis 1999:156). Movies and television frequently are the subjects of these studies (Kidd-Hewitt 1995).

     Many studies focus on the relationship between movie violence and real life behavior. Some studies conclude that people who watch violent movies are more likely than those who watch non-violent movies to exhibit a short term increase in aggressive behavior (Bjorkquist 1985; Black and Bevan 1992). Other research suggests that watching violent movies is an aspect of long term socialization, and produces affects that range from criminal behavior to a desensitization to violence (Blumer and Hauser 1933; Snyder 1995).

     Over the years, researchers in numerous studies have tried to explain the link between violence in (End Page 2) media and violence in the real world. Three explanations seem to be the most prevalent. The first explanation suggests that we learn through observation. This is based on Bandura's (1971) classic social learning theory. A second explanation suggests that the more violence we see, the more we become desensitized to it. The notion here is that, with repeated exposure, we become more comfortable with situations that formerly caused anxiety (Linz, Donnerstein, and Penrod 1988). A third explanation suggests that we see so much violence depicted in the media that we overestimate its presence in the real world, and become unreasonably fearful. Gerbner and Gross (1976) popularized this explanation in their media cultivation thesis, which claims that much of what we know is based upon what we see in the media.

     The most recent comprehensive study of the relationship between media and violence focuses on television. However, the study is relevant to a discussion of movie violence. The National Television Violence Study (1997) is one of the largest, most representative samples of television content ever analyzed. Between October 1994 and June 1995, researchers monitored 23 frequently viewed broadcast and cable channels seven days a week from 6am to 11pm. They conducted a detailed content analysis of 2500 hours of TV programming. Researchers conclude that three problematic effects characterize the depiction of television violence: viewers may learn aggressive attitudes and behaviors; they may be desensitized to violence; they may become fearful (National Television Violence Study 1997:5). These problematic effects concur with the three explanations that we discussed above. Researchers note that, while television violence rarely is graphic, among the most violent (and graphically violent) programs are movies, theatrical releases later aired on premium cable channels (National Television Violence Study 1997:110). The study concludes that TV (including movies shown on TV) depicts a large amount of violence, and that these depictions are highly problematic.

     Recent studies, which reach similar conclusions, fuel critics of the entertainment industry. One study suggests that G-rated animated movies, which are targeted to young audiences, have become more violent over the years, and especially in the 1990s (Yokata and Thompson 2000). Similarly, violence in animated TV shows, a staple of children's programming and very popular among boys, also has increased (Rutenberg 1/28/2001:A1). Some of these cartoons are so graphically violent that the Nickelodeon channel refuses to air them; media critic George Gerbner notes that such violence is so prevalent that Americans are becoming desensitized to it (Rutenberg 1/28/2001:A16). In one last study, The Center for Media and Public Affairs (CMPA), a non-partisan research organization, conducted an exhaustive analysis of 1998's most popular movies, TV series, and music videos. According to the CMPA study, audiences of these popular media saw a serious violent act every four minutes (Fiore 9/23/1999:B8). U.S. Senator Joseph Lieberman said the CMPA study confirms the common belief that "...the culture of violence is alive and killing in America today," and poses a substantial threat to children (Fiore 9/23/1999:B8).

     Thus, concerns about film violence now infuse the political agenda. In 1995, U.S. Senator (and presidential candidate) Bob Dole called movies "nightmares of depravity" (Leland 12/11/1995:46). Columbine and other tragic, high profile incidents that involved "young people killing young people" renewed concerns about violence in film and other media (see Glassner 1999). These concerns intensified when the U.S. Federal Trade Commission (FTC) released the results of a study which condemned the film industry for marketing R-rated movies to youthful audiences (Rosenbaum 9/12/2000:A1,20). FTC Chairman Robert Pitofsky characterized these marketing practices as "deplorable" and a cause for concern (Rosenbaum 9/12/2000:A20). Prompted by the timing of the study, movie violence became an issue in the 2000 presidential election: Democrats and Republicans alike condemned Hollywood for violence in movies, and for marketing violent movies to youth (Rosenbaum 9/12/2000:A1). Politicians from President Clinton to Vice President (and presidential candidate) Al Gore condemned the practices and threatened to sponsor legislation if the movie industry did not stop the abuses (Rosenbaum 9/12/2000:A20). The Republicans even attacked the Democrats for not attacking Hollywood enough: being soft on crime has given way to being soft on violence in the movies as a campaign issue (Perez-Pena 9/21/2000:A22).

     Several interesting points emerge from the FTC study. First, as he condemned the movie industry, Chairman Pitofsky enumerated the same three problematic effects that were noted in the National Television Violence Study (1997), and which generally are used to explain the link between media and real (End Page 3) world violence (Rosenbaum 9/12/2000:A20). Second, even as he enumerated these effects, Chairman Pitofsky acknowledged that academic research has not definitively demonstrated a link between movie violence and violent behavior in the real world (Rosenbaum 9/12/2000:A20). Finally, the newspaper report about the FTC story noted that, according to public opinion surveys, 80 percent of U.S. adults believe that there is a link between violence in the real world and violence depicted in the media (Rosenbaum 9/27/2000:A15).

     What is noteworthy here is that Americans overwhelmingly believe that there is a relationship between movies or other popular media and violent behavior, especially among youth. And yet, the research is somewhat ambivalent. Perhaps human behavior is more complicated than a simple cause (movie messages) and effect (real violence) relationship (see Taylor and Willis 1999). Or, even if such a relationship exists, it may be methodologically difficult to demonstrate given the variety of stimuli to which we are exposed (Hirsch 1981). Or, it may be that the causal arrows go in the other direction: ours is a violent culture, so our movies are violent; indeed, moviegoers develop a taste for movie violence (Gitlin 1991; McKinney 1993).

     Scholars note that Hollywood's pre-occupation with violence is not new. Historically, violence sells (Gitlin 1991; Newman 1993). Even so, there are economic inducements for movies to be more violent today. For example, the violent action in "blockbuster" films needs no translation, which makes their international distribution more lucrative. Profitability is important in the U.S. market as well. Teenage boys, a major demographic target for the film industry, like violent action movies; Hollywood responds (Sparks 1996). The increasing violence in animated television programming also targets boys; there are program tie-ins with action figures which are marketed to boys (Rutenberg 1/28/2001:A16).

     Most scholarly research focuses either on why movies are so violent, or the effects of that violence on real behavior. There are fewer studies which specifically address the increase in movie violence. The studies that address the topic tend to attribute an increase in violence to particular movie genres. For example, Molitor and Sapolsky (1993) analyze "slasher films," a relatively new genre. Molitor and Sapolsky (1993) conclude that violence in "slasher films" increased from an average of 40 violent acts per film in 1980 to an average of 47 acts per film in 1985 to an average of almost 70 violent acts per film in 1989. They add that "slasher films" depict extremely graphic violence, including explicit details of violent attacks and their aftermath (Molitor and Sapolsky 1993).

     Studies also focus on the crime genre and the violent content of such films. Allen, Livingstone, and Reiner (1998) analyze crime films released in England between 1945 and 1991 (most of these films were made in the U.S.). Although the authors find no significant increase in the overall number of crime films (the genre consistently averages about 20 percent of film output in any given year), they do find that the level of violence, and particularly of graphic violence, in the crime genre increases from the 1960s to 1991 (Allen, Livingstone, and Reiner 1998). Similarly, depictions of the effects of violence (trauma to victims) also increase (Allen, Livingstone, and Reiner 1998).

     Pundits and the public alike make two assumptions about movie violence. First, critics assume that movie violence somehow is related to violence in the real world. Second, critics assume that movies are becoming more violent and more graphically violent. In this paper, we address the second assumption: are movies really becoming more violent, and is the violence of an increasingly graphic nature?


     Our objective is to determine if the amount of, and the level of, graphic violence are increasing in Hollywood films. The assumption that "movies are increasingly violent" suggests a (End Page 4) temporal dimension for our research. Accordingly, we analyzed movies over a period of time. In this paper, we analyze one year's worth of films in each decade over a period of four decades.

     As we noted, some researchers focus on the amount and level of violence in specific genres. Although genre research produces useful insights, we are more interested in the violent content of films that are seen by the most people, regardless of genre. Some scholars, including some who focus on genres, use a "most popular film" approach in their research (see Wright 1975). They and we are interested in the films that have "the widest resonance with popular consciousness" (Allen, Livingstone, and Reiner 1998:64). We analyze for violent content the five top-grossing films in one year intervals in each of the past four decades.

     When we first conceptualized the study, the most recent box office data available referenced 1994 films. Similarly, the most recent videos widely available were 1994 films. Thus, 1994 became our end date and, for reasons of symmetry, we chose movies from 1984, 1974, and 1964.

     Initially, we referenced two sources to determine the five top-grossing films in our target years: Film Facts (Steinberg 1978) and Box Office Hits (Sackett 1996). As an aside, the two sources do not always agree in their rankings of the five top-grossing films in a given year. The disagreements are due to each source's decision about the year in which a film should be counted when the film plays in theaters over a two-year period. Because the data for the later decades were incomplete in Film Facts (Steinberg 1978), we relied on Box Office Hits (Sackett 1996) to determine the films for analysis. Here, we encountered a problem not uncommon in film research (the "out-of-print video" problem; see Allen, Livingstone, and Reiner 1998). The Trial of Billy Jack, the number five top-grossing film in 1974 according to Box Office Hits (Sackett 1996), was out-of-print and unavailable in our video rental outlets. As a replacement, we used a film listed in Film Facts (Steinberg 1978) as one of its top five-grossing movies of 1974: The Sting. Indeed, Film Facts (Steinberg 1978) listed The Sting as the number one film of 1974.

     Our next decision involved the matter of a definition of violence. Although thousands of studies examine violence in the media, there is no one commonly accepted definition of violence (see Newman 1998). For our analysis, we use the definition employed in the National Television Violence Study (1997). Violence is "any overt depiction of a credible threat of physical force or the actual use of such force intended to physically harm an animate being or group of beings" (National Television Violence Study 1997). We modified this definition by excluding mere threats of violence from our analysis.

     This modified definition offers several advantages. It is used in the National Television Violence Study (1997), which is one of the most recent and comprehensive studies of violence in the media. It allows us to code actual behaviors, not threatening behaviors, which simplifies coding and calls for fewer interpretations. Finally, the definition includes animate beings as well as people. This was relevant because of cartoons in the television study, and it also becomes relevant for some films in our analysis.

     Next, we defined graphic violence. Our definition is informed by research that addresses graphic violence in terms of blood and gore, trauma to a victim, and realism in these depictions (see Molitor and Sapolsky 1993; Allen, Livingstone, and Reiner 1998). Because critics condemn not just violence but graphic violence, which they claim also is on the rise and which is assumed to be more problematic, we created two definitional categories: one category is for non-graphic violence; the other category is for graphic violence. Category 1 Violence includes those acts that do not depict obvious injury, blood, or trauma realistically. Category 2 Violence includes those acts that depict obvious injury, blood, or trauma realistically. We also wondered if deaths were on the increase, so we counted the deaths that were depicted in these films. Following the same logic that we used in terms of categories of violence, we divided deaths into non-graphic deaths (Category 1 Deaths), and graphic deaths (Category 2 Deaths).

     Our coding instrument was straightforward, if detailed. Following other film research (Molitor and Sapolsky 1993; Allen, Livingstone, and Reiner 1998), we literally counted violent acts. We coded punches, kicks, stabs, gunshots, explosions, or related acts which were directed at animate beings in the movies that we analyzed. In addition to Category 1 and Category 2 Violence and Death, our analysis also includes the (End Page 5) computation of ratios, for example, deaths to violent acts, or graphic violence to all violent acts.

     We offer a coding example for purposes of illustration. Toward the conclusion of Goldfinger, one of the films in our analysis, a scene features a pitched battle between Goldfinger's forces and U.S. military personnel outside of the gold depository at Ft. Knox. When shot, the combatants simply fall down; they could be sleeping. We coded this scene as a Category 1, non-graphic depiction. In contrast, a scene in True Lies exemplifies a Category 2, graphic violence depiction. A man falls from a truck and is run over by the truck. In slow motion, the camera captures the truck's wheels as they run over the man; we see his body twist and contort in pain. The trauma to the human body in this scene is portrayed in a graphic, realistic fashion.

     With subject films, relevant definitions, and a coding instrument in place, we began the analysis. We conducted a close analysis of the films, which necessitated heavy use of the remote control to stop, rewind, replay, and code the violent acts in the films.

     We are satisfied with our method of analysis, although an obvious limitation pertains to the number of films that we reviewed in each year. A deeper cut into each year's top-grossing films say, ten films instead of five might produce different results than those we will present. The five top-grossing films might be an aberration, or they may enjoy some violence-oriented marketing niche that does not persist as one delves deeper into each year's films. Perhaps future research can dig deeper into the list of successful films and reveal more information on this point. In any case, we think that a detailed coding of violence and death in five films per year for four decades yields many insights.

     Our research addresses issues of violence, graphic violence, and whether the movies increasingly depict these acts. We began the analysis with four specific research questions: (1) Are movies becoming more violent? (2) Are movies becoming more graphically violent? (3) Are there more deaths in movies? (4) Are there more graphic deaths in movies?


1964                                                                Gross

Mary Poppins                                               $45,000,000
Goldfinger                                                       23,000,000
The Carpetbaggers                                        15,500,000
My Fair Lady                                                  12,000,000
From Russia With Love                                   9,900,000

     The five top-grossing films of 1964 contain a total of 144 violent acts, or an average of 29 violent acts per film. On closer analysis, however, the films are diverse in their violent content. For example, Mary Poppins and My Fair Lady each depict only four violent acts, while From Russia With Love depicts 70 violent acts. Indeed, just over three-fourths (76 percent) of all violent acts across the five films occur in Goldfinger and From Russia With Love. There are only two acts of graphic violence (one percent of the total) in the five films. Both occur in From Russia With Love.

     The five films in 1964 depict a total of 41 deaths, for an average of eight deaths per film. Two films, Goldfinger and From Russia With Love, account for all of the deaths that are the result of violent acts. Approximately 28 percent of the 144 violent acts result in death. There are no Category 2 deaths (graphically violent) in 1964.

     Most of the violence in 1964 occurs in the two James Bond films. There are only two acts of graphic violence, and they are relatively tame by today's standards, and no graphically violent deaths. The Bond films focus more on depicting a suspenseful or an ingenious way to kill, (End Page 6) rather than on the graphic details of the act.

1974                                                                 Gross

The Sting                                                        $68,500,000
The Towering Inferno                                    52,000,000
Blazing Saddles                                              47,800,000
Young Frankenstein                                      38,800,000
Earthquake                                                     35,800,000

     The five top-grossing films of 1974 contain a total of 168 violent acts, or an average of 34 violent acts per film. These films also are diverse in their violent content. Violent acts are somewhat more evenly distributed across the five films in 1974: Blazing Saddles depicts the most violent acts (76) followed by Earthquake (45 violent acts), The Towering Inferno (24 violent acts), and The Sting (18 violent acts); Young Frankenstein depicts the fewest violent acts (five). The majority of the violent acts (156) are not graphically violent. Twelve acts (seven percent of the total) are graphically violent in the 1974 films.

     The five films depict a total of 44 deaths, or an average of nine deaths per film. Most of the deaths occur in two "disaster" genre films, Earthquake (27 deaths) and The Towering Inferno (10 deaths). Approximately 26 percent of the 168 violent acts result in death. Of the 44 deaths, 38 are in Category 1 (not graphic). Six deaths are depicted in a graphic manner (Category 2).

     The 1974 films depict more violent acts and deaths. There also are more graphically violent acts and graphically violent deaths. The films are becoming more violent, although large amounts of graphic detail are not yet evident.

1984                                                                 Gross

Ghostbusters                                                  $130,200,000
Indiana Jones and the Temple of Doom     109,000,000
Beverly Hills Cop                                           108,000,000
Gremlins                                                             79,500,000
The Karate Kid                                                 43,400,000

     The five top-grossing films of 1984 contain a total of 307 violent acts, or an average of 61 violent acts per film. Once again, the films are diverse in their violent content. Gremlins contains the fewest violent acts (22), while Indiana Jones and the Temple of Doom depicts the most violent acts (141, or 46 percent of the year's total). The Karate Kid depicts 82 violent acts and Ghostbusters and Beverly Hills Cop each depict 31 violent acts. There are 16 acts of graphic violence in 1984 (Category 2). While this is an increase over the two previous time periods, the increase is rather modest given the substantial rise in total violent acts. Indeed, only five percent of the violent acts depict graphic violence.

     Deaths actually decline in the 1984 films. There are 38 deaths, or an average of approximately eight deaths per film. One film, Indiana Jones and the Temple of Doom, depicts half (19) of the deaths. Twelve percent of the violent acts result in death. While the total number of deaths decreases in 1984, more deaths (10, or 26 percent of the total) are graphically depicted (Category 2).

     The films in 1984 depict more violent acts and more graphically violent acts. However, these Category 2 acts decline as a percentage of overall violence. Graphic deaths also increase, although total deaths decline (End Page 7).

1994                                                                   Gross

The Lion King                                                 $298,900,000
Forrest Gump                                                    298,100,000
True Lies                                                             146,300,000
The Santa Clause                                              134,600,000
The Flintstones                                                  130,500,000

     The five top-grossing films of 1994 depict a total of 223 violent acts, or an average of 45 acts per film. Once again, the films are characterized by diversity in their depiction of violence. For example, The Santa Clause depicts only one violent act. Two films, True Lies (111) and Forrest Gump (71), depict the vast majority (82 percent) of the violent acts. The graphic detail of violence (Category 2) increases in 1994. There are 64 acts of graphic violence, and they constitute a larger portion (29 percent) of all violent acts than in the earlier decades.

     The five films depict 76 deaths, or an average of 15 deaths per film. One film, True Lies, accounts for approximately 82 percent of the deaths (62). In 1994, 34 percent of the violent acts result in death. There are 47 graphic deaths (Category 2) in 1994. This represents an increase in terms of raw numbers. Moreover, more than half (62 percent) of the deaths are now of a graphic nature.


     Research Question 1 asks, "Are movies becoming more violent?" The results of our analysis indicate that, yes, overall, in terms of violent acts, the movies in our sample became more violent over the four decades under review. However, this is not a perfectly linear trend: the five films in 1984 depict more violent acts than those in 1994. Otherwise, as we can see in Table 1, in each decade the films that we analyze depict more violent acts than films in the comparable year of the previous decade. And, in any case, the 1994 films depict more violent acts than those in 1964 or in 1974. (End Page 8)

Table 1: Violence Measures of Top 5 Grossing Films by Year (1964-1994)







Total Violent Acts





Average Violent Acts per Film





Distribution of Violent Acts

Mainly 2 films: Goldfinger and From Russia With Love

More evenly distributed

Mainly Indiana Jones and the Temple of Doom

Mainly 2 films: True Lies and Forrest Gump

Category 1 (Nongraphic) Violence





Category 2 (Graphic) Violence





Ratio of Graphic to Total Violence

2:144 = 1%

12:168 = 7%

16:307 = 5%

64:233 = 29%

Total Deaths





Average Deaths per Film





Distribution of Deaths

All in 2 films: Goldfinger and From Russia With Love

Mainly in 2 "disaster" films: Earthquake and The Towering Inferno

Half the deaths in Indiana Jones and the Temple of Doom

Mainly True Lies

Category 1 (Nongraphic) Deaths





Category 2 (Graphic) Deaths





Ratio of Graphic to Total Deaths

0:41 = 0%

6:44 = 14%

10:38 = 26%

47:76 = 62%

     Research Question 2 asks, "Are movies becoming more graphically violent?" As we can see in Table 1, the answer is yes: the number of acts of graphic violence increases in every decade under review. Moreover, with the exception of 1984, the ratio of graphically violent to total acts of violence increases in every decade as well. Again, 1984 is something of an outlier. The 1984 films depict so many acts of violence that, even though there also are more graphically violent acts, the ratio of Category 2 Violence to (End Page 9) all violent acts is smaller than in the films of 1974. A comparison of the beginning and last decades under review is especially revealing. Not only do the acts of graphic violence increase from two (1964) to 64 (1994), but by 1994 more than one-fourth of the violent acts are graphically violent acts (29 percent).

     Research Question 3 asks, "Are there more deaths in the movies?" The answer is, overall, yes, the movies became more deadly in the decades under review. Again, because of 1984, there is not a perfectly linear trend. Indeed, the 1984 films represent an anomaly: far and away the largest number of violent acts, but the fewest deaths. In contrast to the downturn in 1984 (38 deaths), the 1994 films evidence a substantial increase in deaths (76). Moreover, as a ratio (deaths to violent acts), over one-third of all violent acts in the 1994 movies result in death.

     We offer two explanatory comments about the non-linear nature of the films in 1984; both address the mix of films that year. First, as compared with the other years, all five films in our sample from 1984 depict relatively large amounts of violence. For example, even Gremlins, the least violent film in 1984, contains many more acts of violence than the "less violent" films in the other three decades. Second, a young adult film like The Karate Kid or a comedy like Ghostbusters may be violent but, at the same time, less deadly. Thus, the films in our 1984 sample contain many violent acts, but few that end in death.

     Research Question 4 asks, "Are there more graphic deaths in the movies?" In terms of both raw numbers and ratios (Category 1 to Category 2 deaths), the answer is yes, there are more graphic deaths in each decade (see Table 1). Again, the end points tell the story: there were no graphic deaths in the 1964 films, but there were 47 graphically violent deaths in the 1994 films. Moreover, 62 percent of all deaths were graphic deaths (Category 2) in 1994.

     As Table 1 clearly reveals, over the four decades under review, the movies in our sample are increasingly violent and more graphically violent. A caveat is in order, however. Our data also reveal that the depiction of violence and death is not equally distributed across all movies, but rather is usually clustered in one or two films in each period.

     Scholars suggest that some movie genres are more violent than others (Molitor and Sapolsky 1993; Allen, Livingstone, and Reiner 1998). Our analysis agrees. Movies in two genres action and disaster depict the most violence. The two disaster films in 1974, Earthquake and The Towering Inferno, depict 40 percent of all violent acts and more than 80 percent of all violent deaths. Across the four decades under review, the action genre, which includes spy films (see Bennett and Woollacott 1987), is even more violent. These films Goldfinger and From Russia With Love in 1964, Indiana Jones and the Temple of Doom in 1984, and True Lies in 1994 depict a disproportionate amount of the violent and graphically violent acts and the violent deaths which we coded in our analysis.

     Two films in our analysis merit additional comment. Both movies are known as hyphenates; that is, they cross genres (Maltby 1995). Blazing Saddles (1974) is a comedy-western, and Gremlins (1984) is a comedy-horror movie. These hyphenate films, which emphasize humor, depict a substantial amount of violence. Blazing Saddles depicts the most violent acts (44 percent of all violent acts) of the films in 1974; almost one-fourth of all deaths in the 1984 movies occur in Gremlins (the gremlins die). Moreover, some of the action films in our sample also depict violence in ways that are sometimes humorous (see King 1999). Scholars are critical of media presentations that pair violence and humor. The concern is that such a pairing is doubly problematic: it reinforces violence and diminishes its seriousness (National Television Violence Study 1997; also see Zillman and Bryant 1991).

     At the same time, however, we should note that in each period under review, several films appear which depict little or no violence. In 1994, for example, although there is a good deal of (End Page 10) violence and much of it is realistic and graphic, there are three relatively non-violent films. This is an important point because it demonstrates that, notwithstanding the concern about violence in movies, there are very popular indeed, top-grossing movies that depict little or no violence.


     We began this analysis by stating a anomaly: Americans seem to be enamored of popular media and, at the same time, to fear them. Nowhere is this more obvious than with the movies. Box office receipts continue to rise, but Americans blame the movies for a host of social ills. Hollywood even became an issue in the 2000 presidential campaign.

     In this paper, we assessed the criticism that movies are becoming more violent and more graphically violent. We analyzed one year's worth of movies from each of the last four decades. Our findings agree with the conventional wisdom and the scholarly research. The movies in our sample became more violent and depicted more acts of graphic violence. Movies also depicted more deaths, and more graphically violent deaths.

     Of course, we cannot infer from our findings that movies caused the tragedy at Columbine or the other problematic behaviors that critics blame on Hollywood. We did not focus on such "cause and effect" relationships. Todd Gitlin (1994) and other social critics may be correct in the assessment that modern life is more violent and that our popular media, including the movies, reflect that reality. Moreover, as spokespersons for Hollywood note, although movie violence may be increasing, crime, even violent crime, has declined over the past decade.

     This is all to say that the relationship between the movies and real life is more complicated than simple questions of "cause and effect." Yes, as our analysis demonstrates, the movies are increasingly violent. However, rather than simple finger-pointing or Hollywood bashing as a campaign strategy, perhaps other issues should be the focus of future scholarship.

     For example, one point is clear from Columbine and similar incidents: most of the assailants were boys. Perhaps this is due to biology or to child-rearing practices (see King 8/22/1999:A26 for a discussion of playground violence among boys). But, recall the studies about how television markets graphically violent cartoons to boys, and that animated movies increasingly are more violent as well. Again, there is no simple cause and effect relationship but rather, many factors, including culture, which produce both gender and a propensity to violence. In any case, we must pay more attention to the relationship between images of masculinity (including what it means to be a movie hero) and violent behavior. Scholars increasingly include such issues in their research agendas in terms of film (Sparks 1996; Cavender 1999; Rafter 2000) and real life (Messerschmidt 2000).

     A related issue is Hollywood's and our penchant for violent movies. Many scholars note that "violence sells," whether it is newspaper accounts of Jack the Ripper more than a century ago (see Walkowitz 1982; Gitlin 1991) or the latest summer action movie blockbuster. Scholars now address such films and their attraction to movie audiences (King 1999).

     We end on a related and perhaps more upbeat note. Violence is not evenly distributed in the films that we analyzed. Rather, violence, especially graphic violence, clusters in certain movie genres (e.g., action films). This is troubling because action movies are so popular among teenage boys. However, such films also are among the most expensive to produce. There are some indications that action films are slipping a bit in box office appeal. This may be a function of the quality of particular films, or an indicator that the genre is wearing thin. In any case, the box office success of other, less violent films may be a harbinger of shifting sensibility in the audience and in Hollywood (End Page 11) .


* Direct correspondence to Professor Gray Cavender, Arizona State University, School of Justice Studies, Tempe, Arizona, 85257 (E-mail: Gray.Cavender@ASU.EDU).


Allen, Jessica, Sonia Livingstone, and Robert Reiner. 1997. "The Changing Generic Location of Crime in Film: A Content Analysis of Film Synopses, 1945-1991." Journal of Communication 47:89-101.

---. 1998. "True Lies: Changing Images of Crime in Postwar Cinema," European Journal of Communications 13:53-75.

Anderson, Walter Truett. 1990. Reality Isn't What It Used to Be. San Francisco: Harper.

Bailey, Frankie and Donna Hale. 1998. "Popular Culture, Crime, and Justice." In Frankie Bailey and Donna Hale (Eds.), Popular Culture, Crime, and Justice, pp. 1-20. Belmont, CA: Wadsworth Publishing Co.

Bandura, Albert. 1971. Social Learning Theory. New York: General Learning Press.

Beckett, Katherine, and Theodore Sasson. 2000. The Politics of Injustice: Crime and Punishment in America. Thousand Oaks, CA: Pine Forge Press.

Bennett, Tony and J. Woollacott. 1987. Bond and Beyond: The Political Career of a Popular Hero. New York: Methuen.

Bjorkquist, Kaj. 1985. Violent Films, Anxiety and Aggression. Helsinki: Commentationes Scientiarum Socialium.

Black, Stephen and Susan Bevan. 1992. "At the Movies with Buss and Durkee," Aggressive Behavior 18:37-45.

Blumer, Herbert and Philip Hauser. 1933. Movies, Delinquency, and Crime. New York: The MacMillan Company.

Cavender, Gray. 1999. "Detecting Masculinity." In Jeff Ferrell and Neil Websdale (Eds.), Cultural Representations of Crime and Deviance, pp. 157-175. Hawthorne, New York: Aldine de Gruyter.

Clarens, Carlos. 1997. Crime Movies. New York: Da Capo Press.

Fiore, Faye. 1999. "Movies, TV Shows Rife with Violence, Study Says," The Arizona Republic September 23:B8.

Gerbner, George and Larry Gross. 1976. "Living with Television: The Violence Profile," Journal of Communication 26:172-199.

Gitlin, Todd. 1991. "On Thrills and Kills," Dissent (Spring:245-248).

---. 1994. "The Symbolic Crusade Against Media Violence is a Confession of Despair," The Chronicle of Higher Education 23 (February:B5). (End Page 12)

Glassner, Barry. 1999. The Culture of Fear: Why Americans Are Afraid of the Wrong Things. New York: Basic Books.

Gomery, Douglas. 1992. Shared Pleasures: A History of Movie Presentation in the United States. Madison: University of Wisconsin Press.

Gray, Herman. 1995. "Popular Music as a Social Problem: A Social History of Claims Against Popular Music." In Joel Best (Ed.), Issues of Images: Typifying Contemporary Social Problems, pp. 143-158. Hawthorne, NY: Aldine de Gruyter.

Hirsch, P. 1981. "On Not Learning From One's Own Mistakes: A Reanalysis of Gerbner et al.'s Findings on Cultivation Analysis, Part II," Communication Research 8:3-37.

Kidd-Hewitt, David. 1995. "Crime and the Media: A Criminological Perspective." In David Kidd-Hewitt and Richard Osborne (Eds.), Crime and the Media: The Post-Modern Spectacle, pp. 1-24. London: Pluto Press.

King, Dolores. 1999. "U.S. Boys More Violent, Playground Study Says," The Arizona Republic August 22:A26.

King, Neal. 1999. Heroes in Hard Times: Cop Action Movies in the U.S. Philadelphia: Temple University Press.

Krutnik, Frank. 1991. In A Lonely Street: Film Noir, Genre, Masculinity. London: Routledge.

Leland, John. 1995. "Violence, Reel to Real," Newsweek December 11:46-48.

Linz, D., E. Donnerstein, & S. Penrod. 1988. "Effects of Long-term Exposure to Violent and Sexually Degrading Depictions of Women. Journal of Personality and Social Psychology 55:758-768.

Livingstone, Sonia. 1996. "On the Continuing Problem of Media Effects." In J. Curran and M. Gurevitch (Eds.), Mass Media and Society, 2nd Ed. London: Arnold.

Maltby, Richard. 1995. Hollywood Cinema. Cambridge, MA: Blackwell.

McKinney, Devin. 1993. "Violence: The Strong and the Weak," Film Quarterly 46:16-22.

Messerschmidt, James. 2000. Nine Lives: Adolescent Masculinities, The Body, and Violence. Boulder, CO: Westview.

Molitor, Fred and Barry Sapolsky. 1993. "Sex, Violence, and Victimization in Slasher Films," Journal of Broadcasting & Electronic Media 37:233-242.

National Television Violence Study, Vol. 1. 1997. Thousand Oaks, CA: Sage Publications.

Newman, Graeme. 1993. "Batman and Justice: The True Story," Humanity and Society 17:297-317.

---. 1998. "Popular Culture and Violence: Decoding the Violence of Popular Movies." In Frankie Bailey and Donna Hale (Eds.), Popular Culture, Crime & Justice, pp. 40-56. Belmont, CA: Wadsworth Publishing Co. (End Page 13)

Nyberg, Amy Kiste. 1998. "Comic Books and Juvenile Delinquency: A Historical Perspective." In Frankie Bailey and Donna Hale (Eds.), Popular Culture, Crime, and Justice. Belmont, CA: Wadsworth Publishing Co.

Perez-Pena, Richard. 2000. "Lieberman Is Criticized On Violent Entertainment," The New York Times September 21:A22.

Rafter, Nicole. 2000. Shots in the Mirror: Crime Films and Society. New York: Oxford University Press.

Rosenbaum, David. 2000. "Violence in Media Is Aimed At Young, F.T.C. Study Says," The New York Times September 12:A1,20.

---. 2000. "Studios to Curb Marketing of R-Rated Films to Youth," The New York Times September 27:A15.

Rutenberg, Jim. 2001. "Violence Finds a Niche in Children's Cartoons," The New York Times January 28:A1,16.

Sackett, Susan. 1996. Box Office Hits. New York: Billboard Books.

Snyder, Scott. 1995. "Movie Portrayals of Juvenile Delinquency: Part 1 Epidemiology and Criminology," Adolescence 30:53-64.

Sparks, Richard. 1996. "Masculinity and Heroism in Hollywood Blockbusters," British Journal of Criminology 36:348-360.

Steinberg, Cobbett. 1978. Film Facts. New York: Facts on File, Inc.

Surette, Ray. 1998. Media, Crime, and Criminal Justice: Images and Realities, 2nd Ed. Belmont, CA: Wadsworth Publishing Co.

Taylor, Lisa and Andrew Willis. 1999. Media Studies: Texts, Institutions, and Audiences. London: Blackwell Publishers.

Todd, Drew. 2000. "The History of Crime Films." In Nicole Rafter, Shots in the Mirror: Crime Films and Society, pp. 15-45. New York: Oxford University Press.

Walkowitz, Judith. 1982. "Jack the Ripper and the Myth of Male Violence," Feminist Studies 8:543-574.

Wright, Will. 1975. Sixguns and Society. Berkeley: University of California Press.

Yokata, Fumie, and Kimberly Thompson. 2000. "Violence in G-Rated Animated Films", Journal of the American Medical Association (May 24-31) 283:2716-2720.

Zillman, D. and J. Bryant. 1991. "Responding to Comedy: The Sense and Nonsense in Humor." In J. Bryant and D. Zillman (Eds.), Responding to the Screen, pp. 261-279. Hillsdale, NJ: Lawrence Erbaum. (End Page 14)