Hero,Charity Case,and Victim: How U.S. News Media Frame Military Veterans on Twitter

Abstract

Commenters often criticize the mass media for providing audiences a narrow and inaccurate representation of U.S. military veterans. This study examined the claim by researching how regional news publications in the 50 states represented veterans on Twitter. A quantitative content analysis documented the presence or absence of characteristics in 1,460 tweets that employed the terms veteran or veterans. Data were examined using cluster analysis. Three frames emerged. The most prevalent frame, labeled charity, highlighted instances in which veterans received assistance from charitable organizations and others. The second frame, hero, contained references to honor, World War II, and content that would elicit pride from audience members. The third frame, victim, highlighted the mistreatment of veterans by the military and/or society, mental health issues, politics, and the Gulf War. Results suggest U.S. news consumers are provided a narrow representation of what it means to be a veteran.

Keywords

media civil–military relations veterans news stereotype

In the spring of 2017, the Las Vegas Review-Journal shared a story with its readers about how the community would soon be helping homeless veterans “combat life on the streets” (Rogers, 2017). Across the country in Alabama, The Birmingham News website carried an article under the headline, Service Dogs Save Veterans From Themselves (Eversmann, 2017). One month later and one state away, the Atlanta Journal Constitution published two jail mug shots and the headline, Veteran, Soldier Boyfriend Accused of Shooting, Killing Therapy Dog on Video (Ewing, 2017).

The stories covered different topics, appeared in different publications, and reached readers in different states. Nevertheless, the stories shared a common theme: They represented stereotypes concerning veterans of the U.S. Armed Forces, associating military service with mental and physical trauma, increased potential for violence and instability, and a reliance on the charity of others. In isolation, stories about communities recognizing the service of veterans are not problematic. However, veterans and veterans’ organizations say the frequent repetition of stories related to mental illness, violence, charity, and even heroism can be problematic because they provide the civilian public too narrow a portrait of what it means to be a veteran. The National Veterans Foundation (2016) noted that

There is no doubt that serving in the Armed Forces brings about personal and professional challenges, but that does not mean that all veterans are ticking time-bombs…Most of us aren’t standing on a street corner needing a handout.

The mass media, the National Veterans Foundation wrote, are “the main culprit in fostering these negative stereotypes.”

Given such commentary on the mass media, the present study sought to empirically examine how regional news outlets across the United States represent military veterans in stories shared on the popular social media platform Twitter. It is important to understand how news organizations and other mass media represent veterans because the messages conveyed by news outlets, television, film, books, and social media may inform cognitive and affective associations concerning social groups (e.g., Northup & Carpentier, 2015), explicit attitudes toward groups (e.g., Arendt & Northup, 2015), and even behavior related to the group (e.g., Hoewe & Hatemi, 2017). The news media can provide audience members new information and knowledge concerning social groups, especially when news consumers lack firsthand experience and knowledge concerning the group (e.g., Figenschou & Thorbjørnsrud, 2015).

Military veterans represent about 13% of the adult population in the United States, or more than 18.8 million people (U.S. Census Bureau, 2016). Nevertheless, there is little research concerning how the mass media portray this important section of the American population, defined here as people who have served on active duty for any length of time and who were discharged under conditions other than dishonorable. The number of people who closely know a veteran is decreasing (Livingston, 2016), which means the news media can be an important source of information about veterans. The Pew Research Center notes that 8 in 10 Americans between 50 and 64 years old have an immediate family member who served, compared to one in three Americans between ages 18 and 29 (Livingston, 2016). Taking this lack of first-person knowledge into consideration, it is therefore important we understand how the mass media answer the question: What does it mean to be a veteran? The answer could bear significant consequences for both civilians and military personnel, including veterans, given the potential influence of public sentiment on outcomes including governmental policy, veteran reintegration into civilian society, and veteran identity.

In the present study, a quantitative content analysis examined how regional newspapers in each of the 50 states in the United States used Twitter to communicate information about military veterans to the masses. The study contributes to the literature in three ways. First, it goes beyond anecdotal evidence concerning the media’s portrayal of veterans to provide empirical support for the contention that news organizations provide narrow representations of veterans. Second, it analyzes veteran-related stories that come from a geographically diverse sample of news organizations, building on previous research that employed limited sample sizes or examined national publications such as The New York Times and The Washington Post (Kleykamp & Hipes, 2015; Wilbur, 2016). Third, its results provide the foundation for future research into how the mass media shape perceptions of veteran identity.

Literature Review

About 18.8 million military veterans lived in the United States in 2015 (U.S. Census Bureau, 2016). The highest number of veterans served in the Vietnam War (6.8 million), followed by the Gulf War (5.6 million; defined as 1990 to present), peacetime (4.3 million), Korean War (1.8 million), and World War II (WWII; 930,000).

The present study is designed to help us better understand how this important and prevalent social group is framed in U.S. regional news media. Media framing theory proposes that journalists and other communicators shape public perceptions of social issues and actors by focusing on certain bits of information while excluding others (Matthes & Kohring, 2008). Entman (1993, p. 52) defined framing as an action in which communicators “select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation.” To illustrate, commentators criticize journalists for focusing on mental and physical trauma experienced by combat veterans (potentially eliciting pity and/or fear) while affording less attention to the civic and professional leadership potential of veterans who return home (e.g., Phillips, 2015). Veterans might certainly experience trauma. Indeed, about 3.9 million veterans had a service-connected disability in 2015, with 1.2 million of these veterans being rated at 70% or higher¹ (U.S. Census Bureau, 2016). However, by consistently framing veteran experience as traumatic, the media could fail to inform audience members about the potential benefits of service (Kleykamp & Hipes, 2015) and the contributions that veterans can make to their communities upon returning home.

While research concerning this subject is limited in quantity, recent studies have examined how prestigious news outlets frame military veterans and found empirical support for the association between veteran status and victimhood. Kleykamp and Hipes (2015) examined media discourse concerning veterans of Operation Enduring Freedom/Operation Iraqi Freedom in articles from The Washington Post and The New York Times. The articles, published between 2003 and 2011, advanced themes in which veterans were deserving of positive treatment because they were frequently victimized. Sixty-five percent of the articles referenced physical injuries, while 15% of the stories mentioned mental health issues among veterans. Wilbur (2016) examined 10 U.S. daily newspapers and documented similar themes of victimization and charity. Prominent frames included themes of being “broken and disoriented” because of service, a fight against a bureaucratic enemy, an inability of family to cope, and a recommendation that the public help veterans. Coupled with other mass media content, including “wounded hero” films such as Rambo and Brothers, in which veterans commit violent acts because of combat-related trauma, an argument can be made that the American public is being provided a limited portrait of what it means to be a veteran by the media. Veterans have noticed. One veteran and advocate told The New York Times (Phillips, 2015) that the media too often portray a veteran as a:

broken hero, who once did incredible things but is now forever damaged and in need of help. The truth is, 99 percent of us are neither heroic nor broken. We are people—people the public has invested in who have a lot of potential. And it’s time to get over the pity party.

Frames can emerge within news content for several reasons. Shoemaker and Vos (2009) propose that multilevel factors determine whether information makes it through the “gate” and into the news. The ideological leaning of a news organization (e.g., Fox News vs. MSNBC) might inform which stories about veterans are told, and how. Further, simple routines might shape coverage. For example, reporters work under tight deadlines and embrace values such as timeliness and proximity. Therefore, a reporter is likely to cover a community parade on Veteran’s Day, a holiday that centers around themes of sacrifice and honor. Finally, news organizations can be influenced by external forces, such as political actors (e.g., veterans’ organizations or politicians) or social norms concerning military veterans and how we, as society, should treat our service members.

Along those lines, opinion data suggest the U.S. public values military veterans but also endorses stereotypes concerning veterans and demonstrates poor understanding of the lives of military service members and their families (Pew Research Center, 2011a). A study by the Pew Research Center (2011a) found that 9 of the 10 members of the general public who were surveyed took pride in military veterans, and three of the four reported thanking someone for their service. Similarly, participants in a large-scale experiment (n = 4,142; Kleykamp, Hipes, & MacLean, 2017) found that nearly 88% of participants reported they would welcome a veteran moving in next door, while 92% of participants indicated more tax dollars should go toward health care for veterans. In both cases of support, the experiment found social desirability played a role in the responses, although the results still suggested overwhelming support for veterans. Another experiment with 2,892 participants found that the general public is more inclined to stereotype veterans who experienced combat compared to those who stayed in the United States (MacLean & Kleykamp, 2014). However, participants also reported a desire for social closeness and a desire to help veterans, which the authors attributed to the symbolic capital assigned to veterans. While people exhibit stereotypes and prejudice toward people who demonstrate problematic behavior, the status of “military veteran” appears to counteract the inclination and lead to less desire for social distance and the endorsement of help for veterans (MacLean & Kleykamp, 2014).

When it comes to the media’s role in shaping these attitudes, research is scant. However, the existing literature suggests media exposure could indeed shape audience perceptions of veterans. For example, Griffin and Sen (1995) found that participants’ broadcast news exposure influenced who they blamed for problems facing military veterans (e.g., veterans or external forces, such as combat). The results were part of a larger study focused on the most popular fictional movies about the Vietnam War. Griffin and Sen wanted to determine whether the films portrayed veterans in a traditional manner (i.e., “the character controls more of his or her own fate,” e.g., Rambo, p. 516) or situational (i.e., “external forces dominate the character,” e.g., Platoon, p. 516). The researchers then surveyed 355 adults in an attempt to understand how exposure to the movies informed public attribution of responsibility for veterans’ postconflict problems and how attribution informed support for policies benefiting veterans. The more respondents consumed movies in which external forces controlled veterans (e.g., Platoon), the more they blamed external forces for the problems experienced by veterans postconflict. Conversely, the more they watched traditional movies (e.g., Rambo), the less they blamed external forces. In turn, the more people blamed external forces for veterans’ postconflict problems, the more they supported government benefits for Vietnam veterans. Meanwhile, the more respondents blamed the veterans themselves, the more they thought the veterans should help themselves (Griffin & Sen, 1995). The results underscore the potential consequences of media exposure on attitudes and behavioral outcomes such as support for governmental policy.

Media content and public attitudes concerning veterans likely feed off each other. As Wilbur (2016, p. 268) noted, “Since the U.S. military is a product of American society and culture, frames about its veterans will reflect certain shared cultural assumptions.” Before we can investigate the apparent relationship, we must first develop better empirical understanding of how the news media portray military veterans. Our current empirical understanding is limited, in part, because the existing news-related research has studied national publications such as The New York Times and Washington Post, whose audiences are largely well-educated, liberal, and well-to-do financially (Pew Research Center, 2012). In addition to audience demographics, the newsrooms behind these elite publications are generally better endowed in terms of financial resources, time, and staffers than regional news organizations that concentrate on state and local news. This could influence how stories are written and photographed. Finally, the national publications might miss the day-to-day stories concerning veterans that Americans encounter on a regular basis: the story of a veteran returning home, the story of the Veteran’s Day parade, and the story of veterans serving—and leading—their local communities. A number of news values inform whether journalists cover stories, including timeliness, human interest, and proximity. What regional news outlets consider proximate differs from what publications such as The New York Times define as proximate, given the geographical territory the organizations cover. For example, a regional news outlet in Alabama featured a Fourth of July profile of a Huntsville, AL, resident and Army veteran who was known locally for his advocacy work on behalf of veterans and for a red, white, and blue tophat he wore at community events (Haskins, 2016). While the veteran’s community service did appear on the official website of the U.S. Army (Skelley, 2013), it does not appear his work was ever mentioned by The New York Times.

This study seeks to build on previous research by examining how daily publications in all 50 states framed veterans on Twitter, a platform by which news organizations are reaching larger audiences. News organizations generally share headlines, photographs, and short summaries of stories on the social media platform, accompanied by a link to the news organization’s website so that readers interested in reading more about the topic may be directed to the article. Regional news organizations such as the Atlanta Journal Constitution and the Denver Post share content related to local and state events and issues on Twitter, and they also share content related to news events that are gathering national attention. Therefore, the news organizations might share a local story about a veteran returning home from Afghanistan while also sharing stories related to national defense and politics. The fact regional news organizations are increasingly using social media and other online platforms to share content has led for calls (e.g., Hess, 2013) for “local” news to be reconceptualized as “geo social.” Given these considerations, the present research advanced with the understanding that regional news publications, while organized to primarily serve nearby audiences, can essentially reach anyone in the world who has access to a Twitter account.

Twitter is a popular social media platform (Pew Research Center, 2017) that is widely used by news organizations (Pew Research Center, 2011b). In addition, it is used by one in the three U.S. adults between ages 18 and 29 (Greenwood, Perrin, & Duggan, 2016), an age demographic that is seeing less one-on-one contact with military veterans than in previous generations. The age range also captures the majority of service members in the U.S. Armed Forces.

At the time of this study, Twitter permitted users to share 140-character lines of text, called tweets, with other users, called followers. The textual information may be accompanied by photographs, graphs, and other visualizations. The social media platform permits the quick and widespread dissemination of information within society, which is one reason it appeals to working journalists. Indeed, the 50 news organizations examined here had nearly 10.8 million Twitter followers in June 2017, when data were collected. Twitter permits news organizations to connect with broader audiences and demographic groups they might not reach through traditional newspaper and broadcast venues. Further, modern news organizations rarely publish information only in the print product or 10 o’clock broadcast. Rather, the stories they share via Twitter reflect content from the legacy publication (Pew Research Center, 2011b). An examination of how news organizations tweet about military veterans, then, will provide an indication of how the outlet frames veterans across its products. Coding tweets rather than entire articles might certainly lessen the amount of information one gleans from the material. However, the approach is valid for real-world news consumption via social media, as the tweets contain information that would be seen by readers with even passing interest, and headlines often elicit the most attention of news stories.

The study employs a methodological approach that is designed to limit research subjectivity and provide a reliable and valid method for documenting frames within news content (Matthes & Kohring, 2008). Given the limited amount of literature concerning media and veterans, the study advances a research question rather than a hypothesis. It asks:

Research Question: How do regional news media in the United States frame military veterans on Twitter?

Method

Sample

A list of Twitter handles (e.g., @al.com) for the largest circulation publication in each state was compiled, with the exception of California, Virginia, and New York, where the prestige national publications of the Los Angeles Times, The Washington Post, and The New York Times are printed. For these three states, the second largest news publication was chosen. Publications were chosen by cross referencing a marketing firm’s reference list, U.S. Census Bureau data to determine the largest metropolitan areas in each state, and the “number of followers” on the Twitter accounts for news organizations. A list appears in Table 1. Using the advanced search function on Twitter, each publication’s account was searched for tweets that used the terms “veteran” or “veterans.” While alternative words are used to describe people who served in the Armed Forces, we felt the use of these terms would be most encompassing. The search returned a total of 8,417 posts. A random number generator was used to select 20% of tweets from each publication for coding so that the coded sample could generalize to the larger population. Tweets were discarded when they contained the word “veteran” but were not related to the military. For example, a tweet might tell the story of a “veteran athlete” or a “veteran teacher.” In all, coders examined a sample of 1,460 tweets. Dates ranged from April 2008 to June 2017.² The temporal nature of Internet-based content creates challenges for content analysts (Riffe, Lacy, & Fico, 2014). To address the issue, we printed the veteran-related tweets for each news organization (as of June 2017). We also captured screenshots of each account, including information related to the publication’s reach, such as its number of followers (see Table 1).

Table 1.

News Publications and Tweets Referencing Veteran(s).

State	Publication	Tweets Referencing Veteran(s)	Followers
Alabama	al.com	219	244,000
Alaska	Anchorage Daily News	107	63,900
Arizona	azcentral.com	213	326,000
Arkansas	Democrat-Gazette	194	59,900
California	The Mercury News	257	187,000
Colorado	The Denver Post	245	366,000
Connecticut	Hartford Courant	121	133,000
Delaware	Delaware Online	107	95,800
Florida	Tampa Bay Times	193	211,000
Georgia	Atlanta Journal-Constitution	188	850,000
Hawaii	Honolulu Star-Advertiser	170	61,900
Idaho	Idaho Statesman	158	65,400
Illinois	Chicago Tribune	155	929,000
Indiana	Indianapolis Star	189	224,000
Iowa	Des Moines Register	166	134,000
Kansas	Kansas.com	66	54,800
Kentucky	Courier-Journal	120	68,100
Louisiana	The Advocate	88	74,600
Maine	Sun Journal	214	32,400
Maryland	Baltimore Sun	220	265,000
Massachusetts	Boston Globe	298	638,000
Michigan	Detroit Free Press	167	378,000
Minnesota	Minneapolis Star Tribune	178	301,000
Mississippi	The Clarion-Ledger	259	129,000
Missouri	The Kansas City Star	183	214,000
Montana	Billings Gazette	201	28,600
Nebraska	Omaha World-Herald	172	121,000
Nevada	Las Vegas Review-Journal	216	205,000
New Hampshire	Union Leader	123	45,000
New Jersey	The Star-Ledger	200	208,000
New Mexico	Albuquerque Journal	149	61,800
New York	New York Post	121	1,270,000
North Carolina	The Charlotte Observer	131	218,000
North Dakota	Inforum	110	32,400
Ohio	Cleveland.com	299	101,000
Oklahoma	The Oklahoman	95	44,100
Oregon	The Oregonian	184	259,000
Pennsylvania	Philadelphia Inquirer	125	241,000
Rhode Island	Providence Journal	194	96,700
South Carolina	The State	121	83,600
South Dakota	Argus Leader	124	47,000
Tennessee	Tennessean	170	234,000
Texas	Dallas Morning News	183	550,000
Utah	Deseret News	185	86,200
Vermont	Burlington Free Press	105	50,700
Virginia	Richmond Times-Dispatch	112	91,400
Washington	Seattle Times	145	512,000
West Virginia	The Herald-Dispatch	220	12,200
Wisconsin	Journal Sentinel	96	55,600
Wyoming	Casper Star-Tribune	161	8,844
Total		8,417	1,076,894

Procedure

In developing the protocol, we followed a procedure recommended by Matthes and Kohring (2008) for its objectivity, reliability, and validity. Matthes and Kohring recommended breaking down news content into frame elements, which are then coded as present or absent within each piece of news content. The binary variables may then be examined using cluster analysis. The approach helps identify patterns within news content, and these patterns represent media frames (Matthes & Kohring, 2008). More specifically, the analysis places similar items into a cluster and dissimilar items in other clusters. The procedure is now common and its validity supported in media framing research (e.g., Bowe & Hoewe, 2016; David, Atun, Fille, & Monterola, 2011). In the present study, the literature and veteran-related tweets from several news organizations were reviewed in an attempt to compile an exhaustive list of characteristics that might appear in posts related to military veterans. These characteristics appear in Table 2. For the content analysis, coders were asked to examine each tweet and determine whether each characteristic was absent (0) or present (1) within a tweet. In one exception, coders indicated which emotion each tweet would make readers feel toward veterans (happiness, pride, anger, fear, sadness, and surprise). These emotions were later recoded as 0 = absent or 1 = present, producing binary data for cluster analysis. This overall approach treats every news frame as “a specific pattern of variables” (Matthes & Kohring, 2008, p. 264):

Conceived this way, frames are neither identified beforehand nor directly coded with a single variable. Instead, the variables that signify single frame elements are grouped together by hierarchical cluster analysis. The aim of this analysis is the grouping of articles to specific clusters with high differences between the clusters and low differences within a cluster.

Table 2.

Overall Frames and Characteristics Coded Within Content.

Frame	Characteristics Within Frame	Condensed Definition	Overall Prevalence Within Tweets (n = 1,461), Contribution to Clusters
Charity frame	Happiness	The content of the tweet would elicit happiness toward veterans	36% (n = 519; M = 1.00, SD = 0)
	Charity	A veteran or group of veterans receives charity	12% (n = 172; M = .27, SD = .45)
	Physical health	A veteran’s physical health is referenced	6% (n = 91; M = .07, SD = .26)
Hero frame	Pride	The tweet would elicit pride toward veterans from audience members	32% (n = 463; M = .91, SD = .29)
	Honor	The tweet uses the word “honor” in relation to veterans	29% (n = 429 M = .63, SD = .48)
	World War II	The tweet specifically mentions World War II, through visuals or text	10% (n = 140 M = .15, SD = .36)
Victim frame	Sadness	The tweet would elicit sadness toward veterans from audience members	28% (n = 415 M = .96, SD = .19)
	Politics	The tweet references veterans in relation to politics, including veterans engaging with politicians and politicians referencing veterans	15% (n = 213; M = .17, SD = .37)
	Veteran wronged	A veteran is wronged in some way, including suffering because of bureaucracy, victimization through crime	12% (n = 181; M = .39, SD = .49)
	Mental health	The tweet references veterans’ mental health or illness	7% (n = 96; M = .16, SD = .37)
	Gulf Wars	The tweet specifically references the Gulf Wars, through visuals, or text	5% (n = 70; M = .08, SD = .27)
No frame (characteristics appeared in <5% of Tweets)	Vietnam	The tweet makes specific reference to Vietnam War through visuals or text	4% (n = 64)
	Homelessness	A veteran is portrayed as homeless	4% (n = 61)
	Employment	The tweet references veterans in relation to employment training, opportunities, or unemployment	4% (n = 61)
	Korean War	The tweet makes specific reference to Korean War through visuals or text	2% (n = 29)
	Violence toward others	The tweet references a veteran committing violence toward another person(s)	2% (n = 27)
	Violence toward self	The tweet references a veteran committing violence toward himself or herself	1% (n = 20)
	Fear	Tweet would elicit fear toward veterans from audience members	1% (n = 15)
	Anger	Tweet would elicit anger toward veterans from audience members	8% (n = 12)

Before formal analysis began, the primary investigator reviewed the coding protocol with two coders. Coders remained ignorant of the study’s research question and the existing literature until data analysis was concluded to avoid rater bias. Based on the precoding discussion, clarifications were made to the operational definitions for each characteristic in the coding protocol. Next, coders practiced using the protocol on tweets from a publication that was not included in the final study sample. Once they felt comfortable to proceed with formal analysis, coders individually examined a selection of 540 tweets that were randomly drawn from the publications to determine whether the protocol was reliable. Interrater reliability using Cohen’s κ was moderate or strong for all variables used in final analysis, ranging from .60 to .92. Having established the reliability of the protocol, coders proceeded to examine the remainder of the sample. The individual tweet served as the unit of analysis.

Once coders documented the presence and absence of key characteristics for the tweets, a hierarchical cluster analysis was conducted using Ward’s cluster method and a measure of binary/squared Euclidean distance. First, descriptive statistics were used to determine which characteristics appeared at a rate of 5% or more within the data set. Characteristics were excluded when they appeared in fewer than 5% of the tweets because they would not contribute to clustering (Matthes & Kohring, 2008). This reduced the number of characteristics to 11 for use in the analysis, and in the process, it excluded three characteristics that were surprising given the research literature and critical commentary concerning media portrayals of veterans: the veteran as violent, homelessness among veterans, and employment among veterans. These omissions are further addressed in the Discussion section. The hierarchical cluster analysis suggested a three-cluster solution. The number of clusters was determined using three approaches: first, the following heterogeneity measures were examined: 861 (six clusters), 928 (five clusters), 999 (four clusters), 1,111 (three clusters), 1,584 (two clusters), and 2,110 (one cluster). This suggested that merging the third cluster with the second (a difference of 473) would produce a too heterogeneous solution. This conclusion was affirmed by examining a visual representation of the measures, seeking out a clear “elbow” that “signifies that fusing these two clusters would result in a cluster that is too heterogeneous” (Matthes & Kohring, 2008, p. 269). Finally, the three-cluster solution was confirmed using a visual inspection of the dendrogram produced in the analysis. A second analysis using a two-step algorithm was conducted to validate the results of the hierarchical cluster analysis. The same 11 items were entered as inputs. The system automatically determined clusters (with 10 maximum), and Schwarz’s Bayesian criterion served as the clustering criterion. The analysis produced a three-cluster model. Cluster quality, as indicated by silhouette measure of cohesion and separation, was good. Average silhouette width was .5. In addition, the 11 items appeared in the same clusters as they did in the hierarchical analysis.

Results

In isolation, characteristics appearing most often within the tweets included references to honor (appearing in 29% of the tweets; n = 429), instances of charity/assistance toward veterans (appearing in 12% of tweets; n = 172), and situations in which veterans were wronged in some way (appearing in 12% of tweets; n = 181; see Table 2).

The higher the mean, the more important the individual variable was in each cluster. The tweets in the largest cluster, which represented 35% of the overall news tweets about veterans (518 of the 1,460), suggested a charity/assistance frame in which veterans received free goods or services from businesses, health-care providers, fundraisers, and charitable organizations. The frame contained three elements, which appeared in the following order of importance: happiness (M = 1.00, SD = 0), charity (M = .27. SD = .45), and physical health (M = .07, SD = .26). To illustrate, Arizona Central tweeted, “Wounded at war, Army veteran gets new home,” while the Arkansas Democrat-Gazette shared, “Veterans can get free canes at Sam’s Club.” In Jackson, Mississippi, the Clarion-Ledger shared, “Disabled veterans to get property tax break.” The Times-Dispatch in Richmond, Virginia, tweeted “7-Eleven offering veterans a fee-free franchise,” while the Charlotte Observer in North Carolina told its readers that 7-Eleven was “to give away car washes to military personnel, veterans.”

The second (and comparably sized) cluster, referred to as a hero frame, contained three elements: Content that would elicit pride from audience members (M = .91, SD = .29), references to honor (M = .63, SD = .48), and specific mention of WWII (M = .15, SD = .36). The elements appeared in 511 of the 1,460 news tweets, representing 35% of the overall sample. To illustrate, The Argus-Leader in South Dakota informed its readers that “Businesses and organizations nationwide are honoring veterans today.” The Dallas Morning News, meanwhile, informed its readers that, “Dallas photo exhibit focuses on bravery, sacrifice of WWII veterans in senior living center.” In Arkansas, the Democrat-Gazette informed its readers that “70 years later, 94-year-old…veteran awarded 10 medals and distinctions for WWII service.” In one tweet, Burlington Free Press in Vermont wrote, “We offer our greatest gratitude to our veterans and those who are active in the U.S. Military,” while on another occasion, the publication simply tweeted, “Thank you to our Veterans.” Similarly, the Seattle Times in Washington told its readers that “Veterans Day is a day to say thank you,” while the Casper Star-Tribune in Wyoming informed its readers that “This Memorial Day weekend is a time for family and honoring fallen veterans.”

The third cluster, a victim frame, contained five elements: content that would elicit sadness toward veterans (M = .96, SD = .19), a veteran wronged (M = .39, SD = .49), references to politics (M = .17, SD = .37), veterans’ mental health (M = .16, SD = .37), and the Gulf conflicts (M = .08, SD = .27). The cluster included 431 tweets, or 30% of the overall sample. The Boston Globe, for example, shared with its readers the story of a New York town that “demolished a veteran’s house while he was at rehab for surgery.” One day, the Detroit Free Press tweeted, “Suicides of young veterans top those of active-duty troops,” while the publication tweeted on another day, “Veterans struggled to find jobs in Michigan, where unemployment is 2nd in nation.” The Tampa Bay Times in Florida tweeted, “News at noon: Report shows VA staff left veteran’s body in shower nine hours.” Cleveland.com , meanwhile, told its readers in Ohio that “About one out of every 10 veterans could lose access to care under AHCA, according to a new report.”

Discussion

This study sought to better understand how regional news media in the United States framed military veterans on the social media platform Twitter. A quantitative content analysis used cluster analysis to draw out three frames concerning veterans: the hero frame, the charity frame, and the victim frame. The study was conducted at a time when veteran organizations expressed concern about the “wounded hero” model, which combines positive (e.g., heroism) and negative (e.g., charity, traumatized) stereotypes about what it means to be a veteran. Based on the results of this study, those concerns appear warranted. Additional stereotypes appeared in the content, although not often enough to warrant inclusion within the frames, such as stories linking veterans with violent behavior toward themselves and others upon reintegration into civilian society.

Similar to Kleykamp and Hipes (2015) and Wilbur (2016), our quantitative analysis found that news frames often focused on physical and mental trauma, plus the veteran as deserving of charity and gratitude from the public. The frames of charity (the first most prevalent) and victimization (third most prevalent) run contrary to values ingrained in military personnel, including stoicism, physical and mental toughness, and self-reliance (Cooper, Caddick, Godier, Cooper, & Fossey, 2016). The news frames illustrate a divergence between civilian and military cultures, one difference that could make the transition from military to civilian society difficult for some veterans. The frames of victimization and charity also direct attention to the potential negative outcomes of military service. Given the focus, it would be worthwhile to examine how news exposure shapes public attitudes toward serving in the military, including its effect on the attitudes of recruits and their family members. The results suggest the news media might neglect benefits of service and provide audience members a sentiment concerning service that is not shared by the men and women who serve in the Armed Forces. In a survey conducted by the Pew Research Center (2011a), most veterans who served post-9/11 active duty described their service in positive terms, with more than 90% saying the service helped them mature, learn to work well with others, and increased their self-confidence. Post–9/11 veterans reported being happy with their lives, and most said they would recommend military service to a younger person.

The second most prominent frame within the news coverage, which centered around heroism, focused on veterans of WWII. The finding raises questions concerning the passage of time and societal treatment of war and the people who fight. The news stories examined here largely ignored veterans who served during Korea and Vietnam, despite the fact they represent 8.6 million (or 46%) of the veterans living in the United States in 2015 (U.S. Census Bureau, 2016). Instead, the most prominent group that appeared in news content represents the smallest population of veterans still living in the United States. These veterans, who have been labeled “The Greatest Generation,” appeared in this single frame of heroism. The U.S. news media’s focus on WWII veterans reflects public opinion data collected in the United Kingdom, where the public definition for the term “veteran” most often aligns with people who served during the world war (Burdett et al., 2013). This finding bears further investigation given the aging veteran population and the fact that the largest proportion of U.S. veterans will soon be those who served during the Gulf Wars.

It is interesting, too, that certain characteristics did not appear more often in the news content. Three characteristics were excluded from the framing analysis because they each appeared in less than 5% of the content: the veteran as violent, the veteran as homeless, and the veteran as unemployed. These omissions are notable because violence and homelessness are two negative stereotypes often associated with veteran status, and they represent stereotypes that often provide fodder for critiques concerning how the mass media portray veterans. Also, the lack of stories about veterans that involved employment (or unemployment) is interesting. When veterans reintegrate into civilian society, a common concern involves employment. Recent years have witnessed national initiatives to encourage employers to hire veterans and to prepare veterans for the civilian job hunt and workplace. The present study suggests regional news outlets did not often share these stories of employment with readers via Twitter. Veteran advocates and organizations would likely benefit from working directly with local reporters, introducing story ideas and sources who could be used for articles about veterans.

The present research warrants replication in an international context. Different journalistic practices, regulations, and cultural norms shape the work of newsrooms around the world. Following an international survey of 20,000 journalists in 21 countries, Weaver and Willnat (2012) determined there were “more differences than similarities between journalists from these nations, a finding that seemed to reflect social and political differences rather than the influences of media organizations, journalism education, or professional norms” (p. 1). In addition to media practices, countries differ in terms of military service requirements and histories of conflict. We might expect media frames and public attitudes toward veterans to differ in countries such as the United States, Canada, and the United Kingdom, where conscription has ended, compared to South Korea, where men are required to serve 2 years in the Armed Forces. Researchers in the United Kingdom have noted that public attitudes and understanding of service members change when a nation shifts from conscription to an all-volunteer military (Hines, Gribble, Wessely, Dandeker, & Fear, 2015) or vice versa. Military officials and researchers in the United Kingdom have observed the existence of a military/civilian gap similar to the one in the United States partly because of the transition toward all-volunteer personnel, noting that social distance “can arise between the Armed Forces and civilians from a lack of contact and shared experiences, and the implications for mutual understanding and support” (Hines et al., 2015, p. 692). Similar to U.S. residents, the public in the United Kingdom often perceive veterans as hero-victims harmed by violent conflict and government “insensitivity” (Hines et al., 2015, p. 700), characteristics of the “victim” frame documented in our present study of U.S. news stories. It has been noted (McCartney, 2011) that British media often associate military service with heroism and victimization by violence, psychological injury, and government inefficiency, characteristics documented here in U.S. news content. Commonalities appeared in the results of a survey (Mahar et al., 2017) in which public opinion in Canada, the United Kingdom, and the United States looked favorably upon the Armed Forces. Nevertheless, a smaller percentage of Canadian respondents thought military veterans stood greater chance than the general public of being homeless or incarcerated, an outcome the researchers said might be attributable to media coverage. Finally, a nation’s public may endorse different attitudes toward veterans and the Armed Forces based upon the nation’s history of conflict, including the reasons it became involved in conflict (e.g., Berndtsson, Dandeker, & Ydén, 2015 compare British and Swedish public opinion). This is further reason we might expect the news media in two nations to differentially portray the military and veterans.

This study’s results provide the foundation for future work examining the role of media exposure in shaping public attitudes toward veteran-related policy, to the benefit or detriment of veterans. McCartney (2011) highlighted several areas in which stereotypes concerning veterans might influence public policy, including everything from government spending on equipment to overall support for the Armed Forces. Public support for the military could diminish when conflicts elicit negative public sentiment and, in turn, the injuries/deaths of soldiers are perceived as unnecessary. Further, McCartney wrote, the victim stereotype could threaten the success with which the military recruits and retains service members.

Based on the current findings, we might expect people to be more prone to support health-related initiatives after encountering the victim frame in which veterans are portrayed as suffering mental and physical health issues. Given that frames highlight certain bits of information (such as health) while ignoring others (such as employment), it could be that repetitive exposure to this victim frame nurtures public support for health-related policy while diverting attention from the contributions veterans could make within the workforce. Indeed, some scholars question whether the American public would support programs benefiting veterans if the media focused less on deservingness and victimization (see discussion section of Kleykamp & Hipes, 2015). Again, survey data suggest that a relationship might exist between public attitudes and media content. The U.S. public condemns the government’s treatment of military veterans, focusing on mental and physical health treatment. In one survey, 33% of respondents who criticized the government’s support identified general medical problems as an issue, while 22% identified mental health/psychology and 12% identified a lack of jobs and preparation of troops for work (Pew Research Center, 2008).

The present study could be replicated (1) among community newspapers, (2) among publications outside the United States, and (3) via other social media platforms, demonstrating commonalities and differences in the stories told about military service across cultures. In addition, the results presented here raise questions concerning media processes (how media content is produced) and effects (how the media affect audience members). Regarding processes, it is important we understand the reason news organizations employ so few frames to tell stories about military veterans. Gatekeeping theory proposes that a number of factors could be involved, including organizational structure (such as newsroom resources), the attitudes of individual journalists, and social norms governing life in the communities and cultures these news outlets serve. One must also consider the routines of working journalists. The reporters employed by regional publications often write more than one story per day, a constraint that might decrease the likelihood of reporters producing in-depth stories or stepping out of routine. Therefore, a journalist might write about veterans only when service members are making the news because of national holidays or when a person gets into legal trouble and he or she happens to be a veteran. Indeed, a follow-up analysis of this study’s data found that 561 of the 1,460 tweets examined here, or 38%, appeared during May or November, the months in which the United States recognizes Memorial Day and Veterans Day, respectively. By better understanding the processes behind media content, activists could challenge the predominant media frames concerning military veterans. Advocates for veterans could encourage alternative frames, including story lines related to employment and resiliency.

Regarding effects, it is important we understand how these media frames inform public attitudes and behavior concerning veterans. The prevalence of people who have a firsthand connection to a military veteran is expected to decrease as the veteran population ages and moves further into the era of volunteer military service. Further, we need to understand how veterans perceive their own social identity, plus the role of the mass media and public attitudes in shaping those perceptions. The present study represents a foundational step, contributing to our understanding of how print media portray military veterans through analysis of the Twitter posts of 50 regional publications in the United States. The findings reinforce previous analyses, further demonstrating that the U.S. news media provide the general public limited insight into what it means to be a veteran.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research presented in this article was funded by a grant from the Institute for Communication and Information Research, College of Communication and Information Sciences, the University of Alabama.

Notes

References

Arendt

Northup

(2015). Effects of long-term exposure to news stereotypes on implicit and explicit attitudes. International Journal of Communication, 9, 21. doi:10.1027/1864-1105/a000020

Berndtsson

Dandeker

Ydén

(2015). Swedish and British public opinion of the Armed Forces after a decade of war. Armed Forces & Society, 41, 307–328. doi:10.1177/0095327X13516616

Bowe

B. J.

Hoewe

(2016). Night and day: An illustration of framing and moral foundations in the Oklahoma Shariah amendment campaign. Journalism & Mass Communication Quarterly, 93, 967–985. doi:10.1177/1077699016628806

Burdett

Woodhead

Iversen

A. C.

Wessely

Dandeker

Fear

N. T.

(2013). “Are you a veteran?’ Understanding of the term ‘veteran’ among ex-service personnel: A research note. Armed Forces & Society, 39, 751–759. doi:10.1177/0095327X12452033

Cooper

Caddick

Godier

Cooper

Fossey

(2016). Transition from the military into civilian life: An exploration of cultural competence. Armed Forces & Society, 44(1), 156–177.

David

C. C.

Atun

J. M.

Fille

Monterola

(2011). Finding frames: Comparing two methods of frame analysis. Communication Methods and Measures, 5, 329–351. doi:10.1080/19312458.2011.624873

Entman

R. M.

(1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43, 51–58. doi:10.1111/j.1460-2466.1993.tb01304.x

Eversmann

(2017, 3 28). Service dogs save veterans from themselves. al.com. The Birmingham News. Retrieved from http://www.al.com/opinion/index.ssf/2017/03/service_dogs_save_veterans_fro.html

Ewing

(2017, 4 27). Veteran, soldier boyfriend accused of shooting, killing therapy dog on video. Atlanta Journal Constitution. Retrieved from https://www.ajc.com/news/crime–law/veteran-soldier-boyfriend-accused-shooting-killing-therapy-dog-video/KVnhCBIyeSMgFikgJkxcYK/new.html

10.

Figenschou

T. U.

Thorbjørnsrud

(2015). Faces of an invisible population: Human interest framing of irregular immigration news in the United States, France, and Norway. American Behavioral Scientist, 59, 783–801. doi:10.1177/0002764215573256

11.

Greenwood

Perrin

Duggan

(2016, 11 11). Social media update 2016. Retrieved from http://www.pewinternet.org/2016/11/11/social-media-update-2016/

12.

Griffin

R. J.

Sen

(1995). Causal communication: Movie portrayals and audience attributions for Vietnam veterans’ problems. Journalism & Mass Communication Quarterly, 72, 511–524. doi:10.1177/107769909507200303

13.

Haskins

(2016, 7 3). Who’s the guy in the Uncle Sam hat? Meet a great American. al.com . Retrieved from http://www.al.com/opinion/index.ssf/2016/07/whos_the_guy_in_the_uncle_sam.html

14.

Hess

(2013). Breaking boundaries: Recasting the “local” newspaper as “geo-social” news in a digital landscape. Digital Journalism, 1, 48–63. doi:10.1080/21670811.2012.714933

15.

Hines

L. A.

Gribble

Wessely

Dandeker

Fear

N. T.

(2015). Are the armed forces understood and supported by the public? A view from the United Kingdom. Armed Forces & Society, 41, 688–713. doi:10.1177/0095327X14559975

16.

Hoewe

Hatemi

P. K.

(2017). Brand loyalty is influenced by the activation of political orientations. Media Psychology, 20, 428–449. doi:10.1080/15213269.2016.1202839

17.

Kleykamp

Hipes

(2015). Coverage of veterans of the wars in Iraq and Afghanistan in the U.S. media. Sociological Forum, 30, 348–368. doi:10.1111/socf.12166

18.

Kleykamp

Hipes

MacLean

(2017). Who supports U.S. veterans and who exaggerates their support? Armed Forces & Society, 44, 92–115. doi:10.1177/0095327X16682786

19.

Livingston

(2016). Profile of U.S. veterans is changing dramatically as their ranks decline. Pew Research Center. Retrieved from http://www.pewresearch.org/fact-tank/2016/11/11/profile-of-u-s-veterans-is-changing-dramatically-as-their-ranks-decline/

20.

MacLean

Kleykamp

(2014). Coming home: Attitudes toward U.S. veterans returning from Iraq. Social Problems, 61, 131–154. doi:10.1525/sp.2013.12074

21.

Mahar

A. L.

Gribble

Aiken

A. B.

Dandeker

Duffy

Gottfried

… Fear

N. T.

(2017). Public opinion of the armed forces in Canada, U.K., and the U.S. Journal of Military, Veteran and Family Health, 3, 2–3. doi:10.3138/jmvh.3.2.002

22.

Matthes

Kohring

(2008). The content analysis of media frames: Toward improving reliability and validity. Journal of Communication, 58, 258–279. doi:10.1111/j.1460-2466.2008.00384.x

23.

McCartney

(2011). Hero, victim or villain? The public image of the British Soldier and its implications for defense policy. Defense & Security Analysis, 27, 43–54.

24.

National Veterans Foundation. (2016). Stop the negative veteran stereotypes. Retrieved from https://nvf.org/veteran-stereotypes/

25.

Northup

Dillman Carpentier

(2015). Michael Jordan, Michael Vick, or Michael who? Activating stereotypes in a complex media environment. Howard Journal of Communications, 26, 132–152. doi:10.1080/10646175.2015.1011354

26.

Pew Research Center. (2008). Public continues to fault government for troop care. Retrieved from http://www.people-press.org/2008/03/19/public-continues-to-fault-government-for-troop-care/

27.

Pew Research Center. (2011a). War and sacrifice in the post-9/11 era. Retrieved from http://www.pewsocialtrends.org/2011/10/05/war-and-sacrifice-in-the-post-911-era/-executive-summary

28.

Pew Research Center. (2011b). How mainstream media outlets use Twitter. Retrieved from http://www.journalism.org/2011/11/14/how-mainstream-media-outlets-use-twitter/

29.

Pew Research Center. (2012). Demographics and political views of news audiences. Retrieved from http://www.people-press.org/2012/09/27/section-4-demographics-and-political-views-of-news-audiences/

30.

Pew Research Center. (2017, 1 11). Which social media platforms are most popular. Retrieved from http://www.pewinternet.org/chart/which-social-media-platforms-are-most-popular/

31.

Phillips

(2015). Coming home to damaging stereotypes. The New York Times. Retrieved from https://www.nytimes.com/2015/02/06/us/a-veteran-works-to-break-the-broken-hero-stereotype.html

32.

Riffe

Lacy

Fico

(2014). Analyzing media messages: Using quantitative content analysis in research. New York, NY: Routledge.

33.

Rogers

(2017, 3 14). Annual Las Vegas event aims to help homeless veterans get back on their feet. Las Vegas Review-Journal. Retrieved from https://www.reviewjournal.com/news/military/annual-las-vegas-event-aims-to-help-homeless-veterans-get-back-on-their-feet/

34.

Shoemaker

P. J.

Vos

(2009). Gatekeeping theory. New York, NY: Routledge.

35.

Skelley

K. D.

(2013, 11 6). Annual veterans observance grows into week of events. U.S. Army. Retrieved from https://www.army.mil/article/114618/annual_veterans_observance_grows_into_week_of_events

36.

U.S. Census Bureau (2016). FFF: Veterans Day 2016. Retrieved from https://www.census.gov/newsroom/facts-for-features/2016/cb16-ff21.html

37.

Weaver

D. H.

Willnat

(Eds.). (2012). The global journalist in the 21st century. New York, NY: Routledge.

38.

Wilbur

(2016). Defining the Iraq–Afghanistan veteran in American newspapers. Sociology, 6, 267–276. doi:10.17265/2159-5526/2016.04.006