B

Content Analysis of Alcohol-Impaired Driving Stories in the News

Deborah A. Fisher, Ph.D.

INTRODUCTION

Throughout the 30-year period from 1980 to 2010, the United States made substantial gains in reducing alcohol-impaired driving fatalities, which fell from around 24,000 deaths annually in the early 1980s to just over 10,000 in 2010. Along with the number of fatalities, the rate of alcohol involvement in crash fatalities decreased from 60 percent in 1982 to 34 percent in 2010 (Alcoholalert.com, n.d.; NCSA, 2016). This progress came about through the efforts of grassroots community advocacy on alcohol and traffic safety, most notably with the advent of Mothers Against Drunk Driving and the legislative agendas supported at the state and federal levels, including raising the minimum legal drinking age and lowering the illegal per se blood alcohol concentration (BAC) limit for drivers.

Since 2010, however, fatalities and fatality rates associated with alcohol-impaired driving have remained stable. To provide guidance on how to resume progress on the issue, the Committee on Accelerating Progress to Reduce Alcohol-Impaired Driving Fatalities was established and tasked with identifying promising interventions for preventing impaired driving, barriers that exist and approaches to overcome them, and ways interventions may need to be changed or adapted to resume a downward trend.

One area for potentially refining or redirecting prevention approaches is the information environment around alcohol-impaired driving. The ways stories are told in the media (i.e., the slant, the frame, the lens used to view events and issues) have the potential to influence the ways

people think about issues and, therefore, the solutions they believe will be effective and, thus, will support for alleviating problems. The framing of news stories around issues such as poverty, for example, has been shown to affect the way people think about them, with presentations that focus on general outcomes leading to attributions of responsibility to society at large, while those that focus on specific instances of low-income persons align with attributions of individual responsibility (Iyengar, 1990).

Evidence suggests that the way impaired driving is covered in the news can affect individuals’ support for alcohol enforcement and may affect behavior. In a study that manipulated the presence or absence of alcohol as a causative factor in news stories about violent crime, motor vehicle crashes, and other unintended injuries, investigators found that participants who read news stories with alcohol mentions reported greater support for enforcement of alcohol-control laws compared to those who read news stories lacking alcohol mentions (Slater et al., 2012). Yanovitzky (2002) found in a longitudinal study using data from 1978 to 1995 that news coverage of alcohol-impaired driving significantly contributed to impaired-driving-related policy actions, which, in turn, were associated with a reduction in impaired driving among young and high-risk drivers.

The current content analysis of alcohol-impaired driving news stories was conducted to support the work of the committee by providing an assessment of the news environment around alcohol-impaired driving. Content analysis offers a rigorous methodology for qualitatively coding the content of media stories to gather insights into what topics are being presented to the public most often and how that information is being presented or framed (i.e., the context). The results help further our understanding of what messages viewers may be taking away from their interaction with media. Additionally, the results may suggest ways media messaging may need to be modified to support public health efforts.

METHODS

Sampling

As a first step in identifying the universe of news stories from which a sample could be drawn for coding purposes, four newspapers (The Washington Post, The New York Times, USA Today, and The Wall Street Journal) and three cable television news outlets (CNN, Fox News Network, and MSNBC) were selected as sources in an effort to obtain a balance of media outlets generally considered to represent the right, left, and center of the political spectrum. (Initially several other news outlets were targeted, including Politico, The Drudge Report, and Breitbart; however, the news

stories of these outlets were not available as transcripts in LexisNexis and, therefore, they were dropped from the content analysis.)

After these sources were identified, a literature search was conducted in an attempt to gather as comprehensive a universe of stories as possible on alcohol-impaired driving spanning from 2010 through June 2017. The search strategy involved using a series of alcohol terms (e.g., blood alcohol) and driving terms (e.g., drunk driving) appearing in the headline, byline, or appearing at least four times in the story. The exact search strategy was as follows:

This search strategy cast a wide net to ensure that the vast majority of relevant stories were captured and initially led to the identification of 774 articles published across the seven news outlets. Upon further investigation, 156 stories were dropped because on cursory inspection, they did not focus on impaired driving, or full-text access via hyperlinks or transcripts in LexisNexis was not available. The stories eliminated from the sample at this point came from The Wall Street Journal and the three cable television news outlets. This resulted in a sample of 618 news stories for coding. The distribution of news stories across the seven sources is shown in Table B-1.

Coding Protocol

Guided by the work of Shanto Iyengar and other researchers who have conducted media content analyses focused on how issues such as tobacco are framed in the news media (Clegg Smith et al., 2002), a coding scheme was developed and reviewed by the committee members. The coding scheme contained five sections: story descriptors, story frame and

focus, voices represented, values and themes, and story topics. A copy of the final version of the coding scheme is presented after the references section of this paper.

Story Descriptors

Fields completed to provide descriptive information for each story included title, source, date of publication, date of coding, word count, and type of article (e.g., news, editorial, opinion, other).

Story Frame and Focus

A story frame refers to the techniques used to tell a news story that are posited to influence viewers’ attributions of responsibility for social problems and suggest the solutions needed to solve them. As described by Benjamin (2007), an episodic news frame focuses on individual case studies and discrete events; it is likened to taking a telephoto lens in that it gives a close-up picture. In contrast, a thematic frame focuses on trends over time, and highlights contexts and environment; it is likened to taking a wide-angle lens to the issue. Implications of news framing are that episodic frames tend to highlight solutions targeting the individual experiencing the problem; thematic frames, with their broader focus on the surrounding environment and public policies and institutions, highlight solutions aimed at fixing the underlying conditions that cause or contribute to the problem. Applied to alcohol-impaired driving news stories, those that focus on a specific incident of impaired driving (e.g., a crash, an arrest for driving under the influence of alcohol) represent the

application of an episodic frame; those stories that focus on larger societal issues such as through historical and epidemiological pieces or calls for action by communities and officials to reduce the problem represent use of a thematic frame.

For each story about a specific incident of alcohol-impaired driving, several characteristics of the incident were coded, including the following:

• Type of incident: fatal crash, injury crash, property-damage-only crash, arrest without a crash, other;
• Time frame for reporting on the incident: less than 1 week, 1–4 weeks, more than 1 month later, unclear;
• Ramifications reported: physical injuries to victims and perpetrators, psychological impacts on victims and perpetrators, other impacts on victims and perpetrator (e.g., inability to work, financial), impacts on victims’ families, various stages of the adjudication process, other;
• Whether the focus was on the victim versus the perpetrator;
• Whether the incident involved a celebrity or local person of prominence in the community; and
• What substances were reported as being involved in the incident.

For each story that included discussion of broader social or environmental issues, the technique(s) or approach(es) to presenting this wide-angle perspective, the focus, was coded. Coding categories for focus included historical/epidemiological, call for action, program/policy/legislation, alcohol/drinking culture, sales/marketing/promotion of alcohol, court challenges, and so forth. Stories that included discussion of an alcohol-impaired crash or arrest together with broader social issues related to alcohol-impaired driving were coded for both the specific incident (i.e., episodic frame) and the broader discussion (i.e., thematic frame).

Voices Represented

An extensive list of persons whose voices could be represented in stories by being directly quoted, cited, or referenced as sources of information for the story was developed. As with all variables, the list originally developed was expanded throughout the coding process. Voices included law enforcement and judicial officials, victims and their families, perpetrators and their families, attorneys (perpetrator’s, victim’s, and other/legal analyst), traffic safety advocates, government health and safety organizations, alcohol industry or trade association representatives, alcohol retailers, medical/health professionals, and others. Voices were coded for all stories.

Values or Themes

Values or themes represent the underlying message(s) embodied in the story. Examples of themes coded included victimization, injustice, punishment, individual problem/responsibility, collective problem/responsibility, liberty/personal freedom, and so forth. Values may be expressed directly by individuals represented in the story or implied by the type of information or way information is presented in the story. For example, one story reported on the advocacy efforts by the parents of one of two teenage passengers killed when their friend crashed the car they were all riding in after having been drinking at an underage party where the parent of the host joked with partygoers about the alcohol they were bringing into the home. The victim’s parents were quoted calling for more severe penalties for the teen driver and the adult who knew about and permitted the teen drinking party in his home. In addition to the value/theme of punishment, this story was also coded based on the described actions of the adult social host for individual/problem responsibility as well as failure of family, the latter owing to the implied references to irresponsible parenting. Values were coded for all stories.

Story Topics

Story topics are the specific issues related to alcohol-impaired driving that were discussed explicitly in the story. In the coding scheme developed, these 41 topics fell into broad categories:

• Scope of the problem (e.g., societal toll, human toll);
• Responsibility for the problem (e.g., individual, enforcement of alcohol laws);
• Types of prevention or alcohol-impaired driving countermeasures discussed (e.g., prevention/general deterrence, intervention/specific deterrence, harm reduction);
• New tools or technology (e.g., for police, courts, retailers; for drinkers or drivers);
• Community mobilization;
• Impaired-driving system problems/failures (e.g., system overall, court system, alcohol retailers);
• Impaired-driving system successes (e.g., system overall, legislation/policy, alcohol industry); and
• Other topics that emerged during the coding process (e.g., politics/corruption, targeting social/moderate versus heavy/hardcore drinkers, underage drinking, two-tiered justice system).

Story topics were coded for all stories.

Coding Procedures

A single coder experienced in content analysis conducted the coding task during a 5-week period from July 23, 2017, through August 27, 2017. Once the coding process began and stories were read in their entirety, 31 additional stories of the 618 that were originally included in the sample were eliminated, mostly because they were too short and lacked sufficient information for coding (n = 17). This led to a criterion that stories had to be at least three paragraphs in length to provide enough information for coding. Other reasons for eliminating stories were incomplete transcripts or finding that a story did not actually involve impaired driving (e.g., an article on drowsy driving that had one mention of impaired driving as a comparison regarding degree of impairment; a story about how numerous red flags, one of which was impaired driving along with owing back taxes, multiple divorces, and so forth should have resulted in revocation of an individual’s security clearance; a story on research from the Netherlands on effects of alcohol and cannabis on aggression).

Because of the project timeline, the full sample of 587 stories could not be coded. After flipping a coin to randomly start with the first or second story to begin the coding process, every fourth story was coded. If the fourth story was one of the 31 stories eliminated, the next story was coded. This yielded a final sample of 147 coded stories.

As each story was reviewed, in addition to checking the appropriate codes, extensive notes were entered on a hard copy of the coding form. All “other” responses were specified so that emerging issues could be tracked. This led to the addition of several new codes over the course of the task (e.g., pulling justice/judge out of the other category and making it a separate category in voices; adding effects on enforcement/prosecution and targeting social/moderate drinkers versus heavy/hardcore drinkers to values), which are represented in the coding scheme listed at the end of this paper. Given that most of these new codes had emerged about halfway through the coding process, after the full sample of 147 stories was coded, a second pass was made through the first half of the stories to supplement notes taken and see if material needed to be recoded.

Data from the coding forms, excluding handwritten notes, were entered into an electronic form in Survey Gizmo. Data were exported into an Excel file, which was used to develop counts for all data points.

Results

Overall Sample

Of the total sample of 147 coded stories, 139 (94.6 percent) were news focused, 4 (2.7 percent) were editorials, and 4 (2.7 percent) were opinion

pieces. The number of stories in the sample by year of publication were 10 from 2010, 15 from 2011, 16 from 2012, 27 from 2013, 9 from 2014, 33 from 2015, 27 from 2016, and 10 from the first half of 2017.

Other descriptive information for the sample is presented in Table B-2. Seventy percent of the news stories coded came from the newspaper sources, mostly The New York Times and The Washington Post. Newspaper stories tended to be around 550–570 words in length, while cable television news stories were about double that in length. This differential in story length is attributable to the fact that when cable television news outlets cover a story on alcohol-impaired driving, it is often a story that receives national attention. Additionally, depending on the show on which it is televised, the story may be part of an hour-long news magazine show (e.g., The Nancy Grace Show, Primetime Justice, The O’Reilly Factor) in which stories may be discussed in one of three or four show segments or the entire show may be devoted to it.

Newspaper stories, on the other hand, are shorter, due in part to the diversity of types of stories covered in this platform. Newspapers like The New York Times, The Washington Post, and The Wall Street Journal tend to cover both local events in their respective communities and events of national interest. Thus, their coverage of alcohol-impaired driving events includes a combination of highly publicized national incidents as well as short stories of local arrests and crashes, many of which are covered within a day or two of the event’s occurrence before much information is known. These variations in story length between cable television news outlets and newspaper sources may also reflect assumptions about the

TABLE B-2 Frequency Distribution of News Stories Across News Outlets for All Alcohol-Impaired Driving Stories, n = 147

* Transcripts of The Wall Street Journal articles did not provide a word count, and because they were locked, the content could not be copied into a Word file to determine story length.

differential attention spans of the public for material presented in a written versus visual format.

Episodic Versus Thematic Frames

Of the 147 alcohol-impaired driving news stories, 66 (44.9 percent) focused exclusively on discussing a specific incident of alcohol-impaired driving and, thus, are considered to have employed an episodic frame. The remaining 81 (55.1 percent) used a thematic frame, with a discussion of the broader societal issues either alone (e.g., a comparison of different versions of alcohol-impaired driving laws pending in a state’s legislature, a traffic safety organization’s recommendations that all states lower their illegal per se BAC limit from 0.08% to 0.05%) or in combination with an incident of alcohol-impaired driving (e.g., how a fiery school bus crash in a state was the impetus to a shift in norms and attitudes about alcohol-impaired driving and activism that resulted in stronger legislation, or renewed support for a stronger interlock law following the death of a police officer struck by an impaired driver while working on a holiday driving while impaired [DWI] task force).

Research question 1: Whose voices are represented most often and least often in alcohol-impaired driving news stories? Table B-3 presents the frequency distribution of voices that were represented across the 147 news stories. Most stories quoted or referenced information provided by multiple sources so the percentages do not total to 100 percent. The individuals and entities referenced most often in alcohol-impaired driving stories included law enforcement (27.2 percent; often via the public information officer or other official giving details from the police report), traffic safety advocates (21.1 percent), prosecutors (19.7 percent; typically discussing charges filed/to be filed or sentences sought), other attorney or legal analyst (18.4 percent; legal experts not directly involved in the prosecution or defense providing legal analysis and opinions on cases), and victim’s family (15.6 percent; often in the aftermath of fatal crashes such as trial and sentencing of offenders or in stories that follow up on actions prompted by the crash such as new legislation), followed closely by perpetrators (15.0 percent) and the media (15.0 percent; investigative reporters on cable news shows). Those represented least often in alcohol-impaired driving stories were alcohol retailers (2.0 percent), perpetrator’s family members (2.0 percent), victims (2.7 percent), medical and health professionals (3.4 percent), and callers into cable television shows (3.4 percent). The category “other” was represented in 40.8 percent of stories and consists of a diverse array of individuals and entities idiosyncratic to one or two stories, such as the director of public parks and a bicycle

TABLE B-3 Frequency of Voices Represented in Alcohol-Impaired Driving Stories, n = 147

advocacy group representative quoted in a story about an injury crash in which a vehicle driven by a suspected impaired driver entered a biking path and struck a bicyclist, or an official with the National Conference of State Legislators in a story about calls for tougher state impaired driving legislation.

Research question 2: What values or themes are expressed most frequently in alcohol-impaired driving news stories? The frequency with which different values or themes were discussed in the sample of stories is presented in Table B-4. Because many stories were coded as having multiple themes, the percentages do not total 100 percent. The theme of individual problem/responsibility was the most frequently coded value/theme, occurring in half of the stories (50.3 percent). This is consistent with the fact that nearly half of the stories in the sample (44.9 percent) were incident-only stories that detailed an arrest or crash linked to alcohol consumption. In some cases, the perpetrator or others made a direct statement regarding the perpetrator’s responsibility for the crash; however, in most cases, individual problem/responsibility was inferred by the fact that the story provided details of the incident (e.g., offender’s

TABLE B-4 Frequency of Values/Themes Represented in Alcohol-Impaired Driving Stories, n = 147

impairment, driving actions that led to the crash, charges filed/to be filed) with no information suggesting the event was attributable to causes beyond the perpetrator’s decision to drink and drive and subsequent actions on the road.

Other values/themes expressed relatively frequently included other topics associated with crashes and linked to them immediately afterward, such as victimization (28.6 percent), or later at their legal resolution, such as punishment (27.9 percent) and injustice (16.3 percent). These values/-

themes were followed by getting serious or tough on alcohol-impaired driving (13.6 percent) and failure of family (13.6 percent). The failure of family theme resulted, in large part, from the widespread media coverage of the Texas teen, dubbed the “affluenza” teen, who avoided prison for killing four and seriously injuring two persons in an alcohol-impaired driving crash he caused and instead was sentenced to rehabilitation and probation when his defense argued that his parents’ coddling of him and failure to teach him responsibility or consequences for his actions led to his risky decision to drink and drive. Numerous stories about this crash appeared across media outlets in two spates—first in 2013 when the perpetrator’s trial took place and again in 2015 after the perpetrator and his mother fled the United States for Mexico following the posting of a video on social media of an alleged probation violation.

Many values/themes incorporated throughout the coding process appeared with low frequencies (i.e., across six or seven stories), including the protective role of government (4.8 percent), targeting social versus heavy/hardcore drinkers (4.1 percent), avoiding holiday alcohol-impaired driving (4.1 percent), and five other values/themes were expressed even less frequently. Twenty-one stories (14.3 percent) were coded for “other” values/themes that occurred alone or in addition to more commonly expressed values/themes. Examples of these story-specific values/themes include the right to counsel of DWI offenders (in an opinion piece by a trial lawyer), positive-focused media messaging (in a news story on an alcohol-impaired driving campaign developed in one jurisdiction targeting young males), and possible public safety contributions of a ride-hailing service (in a story in which the company linked data on increases in its ridership to decreases in impaired driving).

Research question 3: What specific topics are included in alcohol-impaired driving news stories most and least often? Table B-5 displays the frequency of mentions of a wide array of topic categories and specific story topics across the sample of alcohol-impaired driving stories. The categories of topics represented most often in stories were prevention/intervention approaches to dealing with alcohol-impaired driving (63.9 percent; e.g., prevention/general deterrence, intervention/specific deterrence, sanctions, harm reduction), responsibility for the problem (61.2 percent), scope of the alcohol-impaired driving problem (53.1 percent), and the combination category of emergent issues (72.1 percent). The single-item category community, which consisted of grassroots advocacy efforts, and the category impaired-driving system successes were the categories that were least frequently mentioned, appearing in just 2.0 and 4.1 percent of news stories, respectively.

TABLE B-5 Frequency of Alcohol-Impaired Driving Issues and Specific Topics Represented in Alcohol-Impaired Driving Stories, n = 147

NOTE: Because stories that address an impaired-driving issue or topic category may contain multiple mentions of the subsumed specific topics, the numbers and percentages for topic categories, which are unduplicated counts, and mentions of specific topics are not equal.

Among specific story topics, those with the most mentions were personal responsibility (52.4 percent), human toll (37.4 percent), specific deterrence strategies for reducing recidivism and protecting the public (34.0 percent), sanctions (27.9 percent), and public health toll (23.1 percent); the “other” category of story-specific topics was represented in 46.9 percent of stories. With few exceptions, the remaining specific topics related to alcohol-impaired driving appeared in less than 10 percent of stories. The topics alcohol-impaired driving system successes-alcohol retailer and alcohol-impaired driving system failure/problem-unspecified did not appear in any of the news stories in the sample. Numerous other topics were mentioned just once or twice across the 147 news stories.

Stories with Alcohol-Impaired Driving Incidents (n = 104)

Table B-6 presents the frequency with which news stories involving alcohol-impaired driving incidents involved different types of events (e.g., fatal crash, arrest without crash), substances used, and persons of public interest, including celebrities and local persons of prominence in the community (e.g., local politician, well-known businessperson). In terms of type of incident, more than half of the stories on impaired-driving events were about a fatal crash (56.7 percent) and just over a quarter of stories were about an arrest without a crash (26.9 percent); injury crashes and property-damage-only crashes each accounted for 6.7 percent of stories.

Regarding substance involvement, over 85 percent of stories suggested that alcohol was the sole substance used and the basis for intoxication. It should be noted that most stories of alcohol-impaired driving incidents did not directly specify alcohol involvement, but mentioned that a driver was “intoxicated” or “impaired” and/or that the driver was charged with DUI or DWI. In this overwhelming majority of cases, the assumption was that the substance involved was alcohol; if another drug was mentioned, then it was coded alone or together with alcohol, as described in the story. The combination of alcohol and prescription medications and alcohol and illicit drugs accounted for 3.8 and 2.9 percent of stories, respectively. In four cases (3.8 percent) the driver was reported to have tested positive for all three substances, and in one case (1.0 percent) the driver was speculated to have taken a prescription drug, GHB, alone.

Over the course of the coding process, it was noted that among some stories that appeared in the sample multiple times across sources and over time, reports of substance involvement for the same incident differed. In one crash that appeared in numerous stories in 2013 at the time of trial and again in 2015 after a potential probation violation, most stories referenced alcohol; only a few stories reported alcohol and cannabis

TABLE B-6 Frequency of Incident Types, Substances Involved, and Noted Persons in News Stories with Alcohol-Impaired Driving Incidents, n = 104

or alcohol and Valium and a single story reported all three substances. Some of the discrepancies in substance involvement are likely attributable to the reporting of events within days of their occurrence before toxicology reports are available. In stories that run in one news cycle only, the information on other substances may become available after the story has ceased to be covered. Thus, the involvement of drugs may be underrepresented in news stories.

Celebrities were mentioned in 13.5 percent of stories with alcohol-impaired driving incidents, always as the perpetrator. For local persons of prominence in the community (i.e., those with name recognition and known widely in their community but not celebrities on a national or international scale), such individuals were represented in 15.4 percent of stories as perpetrators and 2.9 percent of stories as victims.

Research question 4: How extensively across time and consequences is the coverage of alcohol-impaired driving incidents in news stories? Information on the distribution of the news stories with alcohol-impaired driving incidents across different time frames and the frequency with which different consequences associated with alcohol-impaired driving events were reported is presented in Table B-7. Nearly two-thirds of stories (63.5 percent) were reported more than a month after their occurrence. This is associated with the fact that 38 of the 104 stories involving alcohol-impaired incidents also discussed broader societal issues, often dealing with advocacy or policy efforts undertaken in the aftermath of a crash. Just over one-quarter of the stories (26.0 percent) reported alcohol-impaired driving events that had occurred within a week.

Regarding consequences stemming from the alcohol-impaired driving incidents, those mentioned most frequently included pretrial processes (e.g., charges, jail, bond, indictment), which were covered in 44.2 percent of stories; victim’s physical injuries, mentioned in 31.7 percent of stories; impacts on victim’s family, found in 26.0 percent of stories; and sentencing of perpetrators, mentioned in 24.0 percent of stories. Again, the “other” category containing idiosyncratic or story-specific ramifications such as a swimmer’s suspension from competition by USA Swimming and loss of corporate endorsements was coded in 55.8 percent of stories. The remaining consequences associated with impaired driving were each mentioned in less than 10 percent of news stories.

Research question 5: To what extent do news stories on alcohol-impaired driving incidents focus on victims compared to perpetrators? Stories on alcohol-impaired driving overwhelmingly focused on perpetrators (78.8 percent), owing in part to the fact that more than half of the stories involved fatal crashes. Although victims were mentioned, stories tended

TABLE B-7 Time Frame of Reporting and Extensiveness of Coverage Over Time of Alcohol-Impaired Driving Incidents, n = 104

to focus on the actions leading up to the incident and consequences for perpetrators in terms of charges, trial, and sentencing. Victims were the focus of stories in 11.5 percent of cases. In three cases (2.9 percent), stories about an alcohol-impaired driving incident focused equally on perpetrator and victim, and in seven cases (6.7 percent), the focus of the story was on another individual other than perpetrator or victim (e.g., a police officer standing trial for perjury for her testimony given in an earlier DWI case who was acquitted of charges; a father who allowed an underage drinking party in his home after which one of the partygoers caused a fatal crash; two teens facing charges for allowing a fellow underage drinking partygoer to drive home).

Stories with a Broader Social/Environmental Focus (n = 81)

Research question 6: What approach to discussing larger societal/environmental issues do news stories taking a thematic frame to the issue of alcohol-impaired driving use? Table B-8 displays the occurrence of different approaches to discussing broader societal/environmental issues across the 81 thematically framed news stories on alcohol-impaired

driving. Because thematic news stories often used multiple techniques to discuss social issues linked to alcohol-impaired driving, percentages do not sum to 100. Approaches or foci used to present a thematically framed story most often involved discussion of program/policy/legislation (39.5 percent), historical/epidemiological trends or information (37.0 percent), and calls for action/more action on the issue of alcohol-impaired driving (19.8 percent). In 17.3 percent of stories on alcohol-impaired driving there was discussion of broader societal issues by discussing alcohol/drinking culture or linking an incident to issues of politics/corruption, bias/preferential treatment by the judicial system, or illegal immigration. The most infrequently used approach to thematic framing was the discussion of sales/marketing/promotion of alcohol (3.7 percent).

Fifty-one unique stories presented a thematic frame through the discussion of a program, policy, and/or legislation; historical context or epidemiological trends; and/or calls to action. To investigate whether the use of these three approaches to thematic framing was associated with discussions about efforts to reduce the occurrence of alcohol-impaired driving through targeting individuals or the general population, the prevention/intervention approaches within specific topics were examined for these 51 stories. The prevention/intervention approaches included three that

are focused on broad populations of all drivers or all drinkers. These include general deterrence strategies (e.g., per se BAC limits, information/educational campaigns relating to impaired driving), harm reduction efforts to separate drinking and driving (e.g., sober rides, use of designated drivers), and other alcohol policies (e.g., alcohol taxes, legislation on dram shop, responsible beverage service, furnishing alcohol to minors). The other two categories of prevention/intervention approaches to alcohol-impaired driving focus on individuals, specifically impaired-driving offenders, and include specific deterrence (i.e., strategies designed to reduce recidivism through means such as ignition interlocks, assessment/treatment, supervised probation) and punishment (e.g., fines, jail).

Across the 51 stories a fairly even distribution across the two foci of prevention and intervention strategies emerged, with 17 stories (33.3 percent) coded as focusing on an individually targeted policy related to specific deterrence or punishment only, 16 stories (31.4 percent) focused exclusively on a general population approach, and 12 stories (23.5 percent) that included a combination of general population and individually focused approaches to dealing with the problem of alcohol-impaired driving. Six stories (11.8 percent) were coded for using a broader social/environmental approach; however, the specific topics discussed were not related to prevention/intervention of impaired driving but issues such as illegal immigration, efforts to establish online criminal registries, and underage drinking education efforts.

Within each of the three framing techniques, there was considerable variability in the patterns of general population versus individual focus of prevention/intervention approaches. For the 32 stories within the program, policy, and legislation category, individually targeted strategies were discussed most often and were contained in 13 stories (40.6 percent), with 9 stories (28.1 percent) addressing strategies targeting the general population, and 7 stories (21.9 percent) focused on both prevention and intervention strategies. For the 16 stories in call to action, the most prevalent approach was a discussion of a combination of general population and individually focused approaches found in 8 stories (50.0 percent), with 4 stories (25.0 percent) discussing individually focused strategies and 2 stories (12.5 percent) discussing general population strategies exclusively. Finally, among the 30 stories employing a discussion of historical context or epidemiological trends, an even distribution was seen, with 10 stories (33.3 percent) focused on general population strategies, followed closely by 9 stories (30.0 percent) focused on individually targeted strategies and 9 stories (30.0 percent) on a combination of general population and individually focused approaches.

DISCUSSION

The current study was undertaken to determine how alcohol-impaired driving is framed in today’s news environment. The sample of 147 news stories was split, with 45 percent using an episodic frame by focusing on a single incident of impaired driving and 55 percent using a thematic frame by placing the issue within a larger social context. Overall, news stories most frequently expressed values of individual problem/responsibility, victimization, and punishment. Within story topics covered, categories represented most often across all stories were prevention/intervention approaches, responsibility, and scope of the problem. Finally, for the entire sample, the voices most often represented in news stories included law enforcement, traffic safety advocates, prosecutors, and other attorneys/legal analysts. In terms of stories with alcohol-impaired driving events, stories overwhelmingly (almost 80 percent of the time) focused on perpetrators.

These findings suggest two trends in alcohol-impaired driving news stories over the past 7.5 years. First, the pattern of codes across these variables is consistent with the preponderance of fatal crashes among stories, which were nearly 57 percent of stories with alcohol-impaired driving incidents or 40 percent of the entire sample of stories. Such incidents evoke more sense of victimization and discussion of punishment than other types of impaired-driving incidents. Additionally, with a focus in stories on the actions of the perpetrator directly leading up to the incident, they communicate a message of personal responsibility for the arrest or crash, in which the individual is held responsible for drinking and driving. Second, the relatively high representation of the category, scope of the problem, and inclusion of traffic safety advocates as the voice with the second highest prevalence across stories reflect the influence of thematic-framed stories dealing with the issue of alcohol-impaired driving in terms of a community problem, with input from advocates who can speak to the statistics on the public health toll and the efforts aimed at reducing the problem.

Although thematic framing of stories places the problem of alcohol-impaired driving within a broader societal context, it does not ensure that approaches discussed for solving the problem necessarily target social and environmental factors. Among the stories using the most frequent techniques for thematic framing, a substantial proportion focused on individually targeted solutions (33 percent) or issues not directly associated with impaired driving (12 percent), often recounting the details of impaired driving events, highlighting individuals’ bad decision making, and discussing the need for action or policy changes focused on efforts to punish offenders or control their drinking and driving. It should be noted that efforts to rehabilitate and control impaired driving offenders are

advanced by traffic safety advocates and popular among policy makers because they not only target offenders but serve to support public health and safety by either keeping offenders off the road or ensuring that they do not drive with alcohol in their systems. Strategies such as ignition interlocks, electronic monitoring of offenders, and intensive supervision and probation allow offenders to remain with their families and keep their jobs while protecting the public and, therefore, are an important part of the arsenal of impaired driving countermeasures.

Focusing on individually targeted policy approaches to impaired driving, however, means that issues such as the culture of alcohol use, sales and promotion of alcohol, and the responsibility of persons and entities other than offenders often receive scant attention in news stories, even those with a thematic frame. Prevention efforts that target socio-cultural and policy contexts are critically important as they hold the promise of creating an environment in which heavy drinking and drinking and driving are not condoned or facilitated and, thus, help prevent the problem from occurring rather than dealing with offenders after an event has taken place. In the current analysis, slightly more than half of the stories in the entire sample used a thematic frame and, among those using the three most frequent thematic framing approaches, only 55 percent included a discussion of solutions targeting the broader social context of impaired driving and population-based approaches to the problem. To move forward in reducing impaired driving fatalities, efforts to engage news outlets in media advocacy on the issue should focus on encouraging news organizations to move away from stories of simple problem descriptions to those that contain more critical analysis of the social/environmental causal factors and discussion of broader-based solutions through primary prevention.

No other content analyses in the published literature were located that matched the focus of this one on alcohol-impaired driving in general or that used a similar sampling frame, with coding of both newspaper and cable television news stories. However, a comparison to another study of framing analysis of impaired driving stories is worth noting. In a study that examined motor vehicle crashes with injuries in U.S. newspapers between 1999 and 2002, researchers reported that among their sample of 473 stories, most reports in newspapers (97 percent) used an episodic frame. In summarizing their results, Rosales and Stallones (2008, p. 477) explained that “the majority of reports included information on the type of crash, but characteristics about people and vehicles were rarely reported.” Thus, according to the authors, newspaper accounts are incomplete sources of information for influencing public attitudes and supporting public health agendas.

Because the current study examined stories involving crashes, arrest-only events, and alcohol-impaired driving without alcohol-impaired driving incidents, a lower percentage of stories were found to have an episodic frame (45 percent), a rate nearly one half of that found in the study by Rosales and Stallones (2008). This suggests that when assessing the news coverage of alcohol-impaired driving, the sampling methods used to select the platforms and the types of news stories to be included in the analysis may substantially affect the results. Stories confined to crashes are likely to recount details of the incident often in its immediate aftermath; stories on impaired driving without a crash event will more likely discuss causal factors beyond individuals and increase the prevalence of thematic framed stories.

The current study provides up-to-date information on alcohol-impaired driving in the news, yet has several limitations. First, the sample size of 147 stories is small. Unlike other content analyses undertaken as part of longer-term research projects, this content analysis had to be planned and executed within 6 to 8 weeks; thus, it was not possible to code a large sample.

Second, the short time frame also made it necessary to have the coding conducted by a single coder rather than a cadre of trained coders. Using multiple coders requires an extensive training and practice phase, ongoing reliability assessments throughout training and actual coding, retraining as needed, and a reconciliation process for the portion of the actual sample of stories that is double coded. Such a process was infeasible in this case, which resulted in a limited sample size and lack of reliability data. However, the fact that the sample was coded in a relatively short period of 5 weeks may have reduced the likelihood of changes in coding practices over time.

Third, because the purpose of the study was to look at the current and recent past years’ news environment, stories had to be coded using archival means, primarily transcripts in LexisNexis. Although transcripts provide a full record of stories developed for dissemination in text format and presented in platforms such as newspapers, transcripts of stories broadcast on television are not as conducive to capturing information as comprehensively. For cable television programs, transcripts of stories can be incomplete for several reasons (e.g., video segments are not seen and the material transcribed is partial with persons in videos often identified simply as “unidentified male” or “unidentified female”). Additionally, because hosts and multiple guests often talk at the same time so that discussion is indecipherable by transcribers, information is missing as places in the transcript will simply designate “cross talk.” Finally, because the coder cannot access the visual content, some material is harder to understand as context is lost. When information is not accessible in its originally

developed format, there are inherent difficulties in coding stories across platforms.

Finally, because the current sample was drawn from a combination of a limited number of newspapers (mostly published on the East Coast) and several cable television news outlets, there may have been selection bias in terms of the types of stories covered in this content analysis. This may have influenced the results in several ways. First, to the extent that there are cultural differences across regions of the United States regarding impaired driving, the values/themes may have reflected those of East Coast and urban areas disproportionately. Second, news outlets that focus on a national audience (such as USA Today and the three cable television news outlets) cover only the most serious or egregious incidents and, thus, do not cover the number and breadth of impaired-driving crashes as news sources that focus on local events in the community or both local and national events (e.g., The Washington Post and The New York Times). The fact that this framing analysis adopted a broad focus on all types of impaired-driving stories not limited to crashes and the inclusion of several national news outlets together likely reduced the occurrence of incident-only (i.e., episodic) stories in this analysis.

In the future, content coding of news stories can be improved in several ways. First, with a longer planning phase, it will be possible to review stories and develop methods specifically tailored to the different story formats. Second, conducting content analyses of current news stories in which content can be recorded in real time in the format it is disseminated (e.g., recording television shows that can be watched by coders) would avoid problems of coding in formats different from that in which stories were originally developed. Third, digital media deserves more attention. Content analyses have already been conducted of posts to online news message boards covering alcohol-impaired driving enforcement, but young people are increasingly turning to nontraditional sources for their news, including online social networks such as Facebook and Twitter, and these are not as easy to access (Connor and Wesolowski, 2009). Use of these platforms and the rise in citizen journalism may have significant implications for what types of content on alcohol-impaired driving are covered as news and how.

REFERENCES

Clegg Smith, K., M. Wakefield, C. Siebel, G. Szczypka, S. J. Slater, Y. Terry-McElrath, S. Emery, and F. J. Chaloupka. 2002. Coding the news: The development of a methodological framework for coding and analyzing newspaper coverage of tobacco issues. ImpacTeen Research Paper Series, No. 21, May.

Connor, S. M., and K. Wesolowoski. 2009. Posts to online news message boards and public discourse surrounding DUI enforcement. Traffic Injury Prevention 10:546–551.

Iyengar, S. 1990. Framing responsibility for political issues: The case of poverty. Political Behavior 12(1):19–40.

NCSA (National Center for Statistics and Analysis). 2016. Traffic safety facts, 2015 data—Alcohol-impaired driving. Washington, DC: National Highway Traffic Safety Administration.

Rosales, M., and L. Stallones. 2008. Coverage of motor vehicle crashes with injuries in U.S. newspapers, 1999–2002. Journal of Safety Research 39(5):477–482.

Slater, M. D., A. F. Hayes, C. E. Goodall, and D. R. Ewoldsen. 2012. Increasing support for alcohol-control enforcement through news coverage of alcohol’s role in injuries and crime. Journal of Studies on Alcohol and Drugs 73:311–315.

Yanovitzky, I. 2002. Effect of news coverage on the prevalence of drunk-driving behavior: Evidence from a longitudinal study. Journal of Studies on Alcohol 63:342–351.

MEDIA CONTENT ANALYSIS:
FRAMING OF ALCOHOL-IMPAIRED DRIVING NEWS STORIES

FRAME AND FOCUS (episodic versus thematic)

1. Alcohol-impaired driving incident featured [if N, skip to Broader Focus]

2. Incident type

   _________ Arrest without crash (i.e., based on driving behavior)

   _________ Injury crash

   _________ Fatal crash

   _________ Property-damage-only crash

   _________ Other

   If Other, specify __________________________

3. Time frame of reporting

   _________ Immediately after incident (i.e., within 6 days)

   _________ Short term (i.e., 1–4 weeks after incident)

   _________ Long term (i.e., more than 1 month after incident)

   _________ Unclear

4. Ramifications of incident covered [Code all that apply]

   _________ Victim’s physical injuries

   _________ Psychological effects on victim

   _________ Other impacts on victim—e.g., lost job/wages, disability

   _________ Impact on victim’s family

   _________ Physical impact on perpetrator—e.g., injuries

   _________ Psychological effects on perpetrator—e.g., guilt, remorse

   _________ Other impacts on perpetrator—e.g., financial, job

   _________ Pretrial—e.g., charges, bond, home confinement

   _________ Trial

4. _________ Sentencing

   _________ Other

   If Other, specify __________________________

5. Perpetrator–Victim Focus

   [Code both Perpetrator and Victim only if 50-50 split in focus]

   _________ Perpetrator

   _________ Victim

   _________ Other

   If Other, specify __________________________

6. Celebrity—involvement of national/international public figure (e.g., entertainer, athlete) [Code all that apply]

   _________ Perpetrator

   _________ Victim

   _________ Passenger in vehicle with suspected impaired driver

   _________ Passenger in vehicle with victim

   _________ Other (e.g., party host)

   If Other, specify __________________________

7. Local person of prominence/interest—involvement of local person known in the community before incident such as high-profile businessperson, elected official [Code all that apply]

   _________ Perpetrator

   _________ Victim

   _________ Passenger in vehicle with suspected impaired driver

   _________ Passenger in vehicle with victim

   _________ Other (e.g., party host)

   If Other, specify __________________________

8. Substance involvement in incident [Code all that apply]

   _________ Alcohol

   _________ Rx drugs

   _________ Illicit drugs

   _________ Drugs, unspecified

9. Broader focus on societal/environmental issues [Code all that apply; If N, skip to Voices]

   _________ Historical/epidemiological

   _________ Call for action/more action on alcohol-impaired driving

   _________ Program/policy/legislation

9. _________ Alcohol/drinking culture

   _________ Sales/marketing/promotion of alcohol

   _________ Court challenges to laws, police practices

   Links to other social issues

   _________ Politics/corruption

   _________ Two-tiered justice system

   _________ Illegal immigration

   _________ Other

   If Other, specify __________________________

   VOICES (all stories)

10. Whose voice is represented by providing information, being quoted or cited as source of info [Code all that apply]

    _________ Law enforcement

    _________ Prosecutor

    _________ Policy maker

    _________ Victim

    _________ Victim’s family

    _________ Perpetrator

    _________ Perpetrator’s family

    _________ Traffic safety advocates—e.g., MADD, IIHS, community alcohol-impaired driving task force/coalition

    _________ Government public health/safety org.—e.g., NHTSA, NTSB, CDC

    _________ Alcohol industry or industry-funded lobbying/advocacy groups

    _________ Alcohol retailers—e.g., owners/managers of local bars, clubs, or restaurants

    _________ Researcher

    _________ Medical/health professional

    _________ Perpetrator’s attorney

    _________ Other attorney/legal analyst

    _________ Media

    _________ Callers into show

    _________ Other

    If Other, specify __________________________

VALUES (all stories) [Record type of content included in categories]

11. What are the underlying values/themes expressed in the story [Code all that apply]

    _________ Victimization

    _________ Injustice

    _________ Punishment

    _________ Individual problem or responsibility

    _________ Collective problem or responsibility

    _________ Alcohol retailer/industry problem or responsibility

    _________ Failure of family

    _________ Entitlement/above the law

    _________ Liberty/personal freedom

    _________ Effects on alcohol-impaired driving enforcement/prosecution

    _________ Economic interests

    _________ Cultural change in norms/attitudes

    _________ Personal engagement/activism

    _________ Protective role of government

    _________ Influence of special interests beyond industry

    _________ Recovery/personal transformation

    _________ Targeting social versus heavy/hardcore drinkers

    _________ Need for evidence-based prevention

    _________ Experiential learning to prevent risky driving

    _________ Avoiding holiday alcohol-impaired driving

    _________ Getting serious/tough on alcohol-impaired driving

    _________ Other

    If Other, specify __________________________

    TOPICS (all stories)

12. What topics are discussed in the story [Code all that apply]

    Scope of the Problem

    _________ Public health toll—effects on society in terms of deaths, injuries, costs

    _________ Human toll—impact on victims, survivors, offenders

    Responsibility

    _________ Personal—responsibility of individual drinkers, parents of minors, other

    _________ Third party-enforcement (e.g., ABC, police)

    _________ Third party-court and/or probation system

1. _________ Third party-government/policy makers

   _________ Third party-alcohol retailers

   _________ Third party-alcohol industry

   _________ Third party-other

   Prevention/Intervention Approach

   _________ Prevention/general deterrence—countermeasures directed at general driving population (e.g., per se BAC limits, public information/education, checkpoints)

   _________ Intervention/specific deterrence—countermeasures directed at offenders to reduce recidivism and support public safety (e.g., license suspension/revocation, ignition interlocks, DWI courts/ISP, assessment/treatment services, court-mandated sobriety)

   _________ Sanctions—countermeasures focused on punishment of offenders (e.g., fines, jail, home confinement)

   _________ Harm reduction—efforts to separate drinking and driving (e.g., sober ride programs, designated driver, ridesharing [Uber, Lyft])

   _________ Other alcohol policies (e.g., taxes, restrictions on use/density, dram shop, RBS, alcohol promotion)

   Tools/Technology

   _________ New tools/technology for police, courts/probation, retailers—geospatial mapping of crashes/calls for service, passive sensors, tools for checking IDs, etc.

   _________ New tools/technology for drinkers/drivers—BAC estimators, breathalyzers in cars, driverless cars, etc.

   Community

   _________ Grassroots advocacy efforts—community-based efforts (e.g., alcohol-impaired driving coalition/task force, work of traffic safety advocates, other citizen-led efforts)

   Alcohol-Impaired Driving System Failures/Problems

   _________ Failures in alcohol-impaired driving system overall/unspecified

   _________ Failures in/greater action needed on alcohol-impaired driving legislation/policy

   _________ Failures in/greater action needed on enforcement of alcohol laws (e.g., sobriety checkpoints, compliance checks)

   _________ Failures in/greater action needed from court system (e.g., sentencing offenders, monitoring offenders)

   _________ Failures in/greater action needed from alcohol retailers (e.g., RBS training, improved ID checking, dram shop liability)

   _________ Failures in/greater action needed from alcohol industry

1. _________ Failures in/greater action needed from others (e.g., universities, fraternities)

   If Other, specify__________________________

   Alcohol-Impaired Driving System Success

   _________ Alcohol-impaired driving system overall/unspecified

   _________ Successes on alcohol-impaired driving legislation/policy

   _________ Successes on enforcement of alcohol laws

   _________ Successes related to court system (i.e., judiciary, probation)

   _________ Successes related to alcohol retailers (e.g., RBS training, improved ID checking)

   _________ Successes related to alcohol industry

   _________ Successes related to other entities (e.g., universities, fraternities)

   If Other, specify __________________________

   Emergent Issues

   _________ Politics/corruption

   _________ Entitlement/above the law

   _________ Two-tiered justice system

   _________ Irresponsible parenting/family dysfunction

   _________ Targeting social versus heavy/hardcore drinkers

   _________ Change in lifestyle/direction/purpose

   _________ Underage drinking

   _________ Individual vs. state’s rights

   _________ Illegal immigration

   _________ Other issues (e.g., holiday drinking, gang violence)

   If Other, specify __________________________

NOTES