4

TALKING ABOUT POVERTY:

NEWS FRAMING OF WHO

Is RESPONSIBLE FOR CAUSING

AND FIXING THE PROBLEM

By Sei-Hill Kim, John P. Carvalho, and Andrew G. Davis

This study explores how the American news media fiame the poverty issue, looking at the way the media present its causes and solutions. It also examines the notion of fiame building, exploring the factors that may infiuence the media's selective uses of certain frames. Media attribu- tions of responsibility are largely societal, focusing on the causes and solutions at the social rather than personal level. Liberal newspapers made more references than conservative papers to social causes and solu- tions. Television news is slightly less likely than newspapers to make social-level attributions.

With the two recessions in the last ten years, poverty has been on the rise, from abouf 11% in 2000 to over 13% in 2008.' Since the financial crisis of 2008, in particular, renewed attention has been paid to the issue, producing an increasing amount of public and privafe discussion. At the center of fhe discussion is the question of who is responsible for causing and fixing fhe problem. How fo define responsibilify is imporfanf because if may shape fhe overall policy approach, particularly the domain of sociefy fo which fhe change effort should be applied.^

News media play an important role in the process of defining a social problem. The media fiame an issue in a cerfain way, telling the audience what is important to know about and how to think about it."* In particular, Iyengar has demonstrated that news media can frame quesfions of responsibilify, leading the audience to determine the causes of and solufions for social problems.* Analyzing newspaper articles and television news, this study examines how the American news media have presented poverty in their news coverage. More specifically, if looks af how fhe media frame fhe question of who is responsible.

Many sfudies of news framing have examined, in a largely descripfive way, how the media present a certain issue.'̂ Some have test- ed effecfs on fhe audience, looking af different types of framing effects, including perceived issue importance" and evaluations of election candi-

Sei-Hill Kim is an associate professor in the School of Journalism and Mass Communi- J&MC Quarterly cations at the University of South Carolina. John P. Carvalho is an associate professor Vol. 87, Nos. .V4 in the Department of Communication and Journalism at Auburn University, where AutumnlWinter 20W

©2010 AEJMC Andrew G. Davis is an instructor.

TALKING ABOUT POVERTY 563

dates.' Left largely unexplored, as Scheufele poinfs ouf, is fhe question of what makes the media fiame an issue in a certain way.* Drawing upon the sfudies of agenda building, our study explores the nofion oí frame building," which addresses the question of whaf factors may influence fhe media's selective uses of cerfain frames. More specifically, we examine a few macro- and micro-level factors of frame building, looking at how fhese facfors interplay wifh the framing process, shaping fhe way the media frame the poverfy issue.

Frame While agenda-building sfudies'" deal with factors fhat may Building influence fhe media's selecfion of issues to report,/rame building address-

es how the media select specific frames in presenting the issues." The term frame building captures whaf roles are played by social and structur- al facfors in the media system and by the characteristics of individual journalisfs in influencing fhe producfion and modificafion of frames.

Based upon fhe findings from agenda-building sfudies,'^ Scheufele laid out a number of internal and external factors of news organizations thaf may affect how journalists frame a given issue." First, social norms and cultural values can affecf the way an issue is framed. The media fend fo portray society as fundamentally sound, affribufing mosf social prob- lems fo irresponsible or unfortunafe individuals.'* Problems are consid- ered as personal in nature and disassociated from larger social factors. Supporting the argument, Kim and Willis found fhaf news coverage of obesify fended to focus on individual, rather than sociefal, causes and solutions largely because the personal-level attributions of responsibilify were befter fiffed to the strong individualism ingrained in American cul- fure."

Organizational pressures and constraints are another factor. The over- all polifical orienfafion or fhe view of the publisher can be one form of organizational pressure (or norm), offen reflecfed in the ediforial fone or fhe organizafional roufine of each news organization.'* Commercial pres- sures can be anofher form. News organizafions are for-profif organiza- tions; fhe items that attract larger audiences become a regular choice. Likewise, framing a story in a way fhaf attracts large audiences may become necessary when journalisfs consider which frames fo use. Kim, Carvalho, Davis, and Mullins, for example, found that a large majority of news coverage of illegal immigration framed the problem as a crime issue.'' Linking illegal immigration fo a crime sfory must be a preferred way of framing the issue because if will necessarily involve drama, con- flict, good, and evil—the ingredients to attract a larger audience.

Pressures from interest groups also can be considered as a frame- building facfor. Interest groups, according to Edelman, "categorize beliefs in a way fhat marshals support and opposition fo fheir inter- ests.""* In an effort fo consfruct public opinion, interest groups seek to use the media as fools fo establish certain frames of reference. Frames advo- cafed by inferesf groups as sound bifes are offen adopfed by journalisfs and incorporated into their news coverage." As Gamson and Modigliani posit, the media's selection of specific frames can be accounted for by

564 JouRNAUSM & MASS COMMUNICATION QUARTERLY

the interplay between journalistic norms and the influences of interest groups.™

The way news is framed can be an outcome of professional routines of journalists.^' Reliance on routine sources, such as government offi- cials, press releases, and press conferences, is typical.^- Another routine is the frequent use of episodic framing. Episodic framing involves story- telling, in which a topic is presented in a specific event or in a personal case. Thematic fiaming, on the other hand, places a topic in a larger and more abstract social context. While episodic framing is relatively easier to prepare, thematic framing requires a significant amount of back- ground research and data collection. For journalists operating under tight time constraints, episodic framing is the preferred way to prepare a story.^' Episodic framing is also preferred because it often includes human interest stories, an easier way to attract large audiences.

Finally, and perhaps most important, the characteristics of individual journalists can play an important role in deciding how to frame a story.-** Selection of a specific frame is moderated by such factors as social and political ideologies, attitudes, and professional norms, the factors that are often reflected in the way journalists make sense of the issue and construct a story.^'

News media seek to reduce the complexity of the issues by pre- tramtng senting them in easy-to-understand interpretive packages. Framing Responsi- serves as a packaging process, which involves selecting certain aspects bilitv of reality and making them more salient, while leaving other aspects out of the package.̂ ** It is in this selection process that the media promote a particular problem definition, leading the audience to make attributions of responsibility."

Attributions of responsibility can be categorized into two types: causal and treatment responsibilities.^* Causal responsibility deals with the source of the problem, while treatment responsibility focuses on who has the power or the responsibility to alleviate the problem. More simply put, causal responsibility addresses the question of who causes the problem, while treatment responsibility identifies who should be held responsible for solving it. These two definitions of responsibility are particularly useful in understanding public dialogues on social problems. Social problems are inherently political in nature, involving compromise and conciliation of competing interests.̂ '* It is critically important for these competing interests to be able to successfully define the responsibility because it will eventually determine which interest (e.g., the government, industry, or individual citizens) should be held responsible for fixing the problem.'"

The discussion of responsibility involves two conflicting views." One view holds that a social problem is caused largely by the deficien- cies of individuals, often those who are affected by the problem. Because the problem is considered as resulting from flaws in individual behaviors, change efforts tend to focus on modifications of the problem- atic behaviors. According to the other view, a social problem results pri-

TALKINC ABOUT POVERTY 565

marily from flaws in social and environmental conditions, such as unequal distribution of economic resources, unsafe environment.s, or unethical business practices. Remedies require changes in government policies, business pracfices, and/or other larger social forces.'̂

News media are often criticized for reducing important social issues to mere individual-level problems." Social responsibilities are largely ignored, while individual causes and solutions are repeatedly empha- sized. There are several reasons. First, as mentioned earher, the emphasis on individual responsibility may well reflect the strong individualism of the American culture.'"' Second, societal approaches are often considered too radical either politically or economically, and are thus rejected in news coverage in favor of less drastic measures, like changing individual behaviors.''* Prescriptions for such "societal remedies" include, for exam- ple, new laws, new taxes, and new regulafions. Finally, news coverage of social problems is typically episodic, focusing largely on a specific event that has happened to certain individuals.'* It is in this act of presenting a specific event that the media necessarily divert attention from social con- ditions and instead focus more on individual accountabilifies.'^

While some researchers assume that the media play an acfive role in shaping people's understanding of who is responsible,'" a recent study by Kensicki brings into question the media's role." Her content analysis of news arficles about poverty, pollufion, and incarcerafion indicated that news coverage was overwhelmingly "neutral," and rarely suggested any causes, effects, or responsible agent for the problems. Through the over- whelming uses of neutral frames, she concludes, newspapers have pro- moted general political apathy among readers, creating a disconnection between the problem and actual ramifications.

Research Questions and Hypotheses

Our research quesfion (RQ) and the first hypothesis explore how the media frame the question of who is responsible for poverty:

RQ: How have the media presented the causes and solu- fions to poverty? Have certain causes and solutions appeared more often than others?

HI: News coverage of poverty will be more likely to attribute responsibility to individuals than to society.

It can be hypothesized that the overall polifical orientafions of news organizafions (i.e., the view of the publisher) affect the frame building process. Conservafism, for example, emphasizes individual freedom and responsibility and advocates limited government regulations. Conservafive papers, therefore, may likely define poverty as an individ- ual matter. Liberal papers, on the other hand, will be more likely to call for greater government involvement and societal approaches.

H2a: Conservative papers will be more likely than liberal papers to attribute responsibility to individuals.

566 ASS COMMUNICATION QUARTERLY

H2b: Liberal papers will be more likely than conserva- tive papers to attribute responsibility to society.

Compared to newspapers, television is distincfively episodic and event-oriented."" By presenting news episodically, television necessari- ly focuses on what happened to an individual, displacing attention to larger social conditions. It is likely that the poor are frequently seen as being responsible for causing and solving fheir own financial hardship; at the same fime, attenfion is diverted from pofential sysfemic or insti- tutional causes of suffering."' We hypothesize that the emphasis on personal-level responsibility would be greater in television than in newspapers:

H3a: Television news will be more likely than news- papers to attribute responsibility to individuals.

H3b: Newspapers will be more likely than television news to attribute responsibility to society.

Sample. In order to select the newspapers to analyze, we first Methods selected four states from the top ten in median household income (Minnesota, Massachusetts, Virginia, Colorado) and four states from the bottom ten (New Mexico, Oklahoma, Alabama, West Virginia), based on a three-year-average (2004-2006)."̂ In selecting these states, we also tried to include as many different regions of the country as possi- ble. We then selected one newspaper from each sfate, based on its history, circulation size, and availability in the LexisNexis database: Star Tribune (Minnesota), Boston Globe (Massachusetts), Richmond Times Dispatch (Virginia), Denver Post (Colorado), the Tulsa World (Oklahoma), Birmingham Neivs (Alabama), Gharleston Gazette (West Virginia). Because New Mexican newspapers had a relatively small number of poverfy arficles, we included two newspapers from that state—instead of one—(the Santa Fe New Mexican, Albuquerque Journal). The Tulsa World, Birmingham News, Albuquerque Journal, and Richmond Times Dispatch endorsed the Republican candidate in the past three presidential elections, and thus were categorized as conservative newspapers. The Gharleston Gazette, Santa Fe Nezv Mexican, Boston Globe, and Star Tribune endorsed the Democratic candidate in the past presidential elections, and thus were categorized as liberal newspapers. The Denver Post endorsed Gore in 2000, Bush in 2004, and Obama in 2008, and also was categorized as a liberal paper."' As far as television news, we selected three broadcasting networks (ABC, NBC, CBS) and one cable news channel (CNN).

Our analysis included news stories published or aired befween January 1, 1993, and December 31, 2007. The years 1993 to 2000 repre- sent the tenure of the Clinton administration, during which the econo- my was showing continued growth. During the remaining seven years (the Bush administration, 2001 to 2007), on the other hand, the economy

TALKING ABOUT POVERTY ^

stmggled with fwo recessions. The fiffeen-year fime period was selecfed in order to include years when the economy was strong and when it was stmggling.

The actual news data for the study came from a keyword search of fhe LexisNexis dafabase.** Using fhe keyword "poverty" appearing in the headline, we retrieved stories about poverty in the nine newspapers and four felevision channels. The search yielded a fofal of 1,028 news arficles and 402 franscripfs. By checking through the headline of each sfory, we firsf sorfed ouf unrelated items, including stories about poverty in other counfries and duplicates (the same article appearing in different weekly édifions or as a fease for lafer in fhe program). Abouf 53.9% {N = 771) were unrelated items and thus eliminated in this first-round filtering. For fhe newspapers and felevision channels fhaf yielded more fhan sixfy sfories affer fhe filfering, we randomly selecfed sixfy fo analyze. If fhe fil- fered sfories numbered fewer fhan sixfy, we analyzed fhem all. As our analysis proceeded, we were able to furfher sort ouf a small number of unrelafed ifems. The final number of ifems analyzed was 506 (395 articles and 111 franscripfs).

Coding. Table 1 illusfrafes how fhe coding insfrument specified what may constitute each potential cause or solution to poverty. Attributions of causal responsibility were categorized into personal and societal causes. Personal-level causes included individual behaviors, lifestyles, and other factors that might be responsible for poverty. These personal attributions were categorized into one of five causes: Making a bad choice. Broken family. Lack of education. Physical conditions, and Immigration (see Table 1 for details). Societal-level causes, on the other hand, were the social, economic, and political factors fhaf mighf con- fribufe fo fhe poverfy problem. These societal reasons were also cafego- rized into five causes: Bad economy. Inadequate pay. Ineffective government aid. Lack of education support, and Racism.

Attributions of solution responsibility were also categorized into personal and societal solutions (see Table 1). Personal-level solutions included fwo categories, including Making better choices and Education. Societal-level solutions had six categories: Better economic policies. Higher minimum wage. Reforming government aid programs. Greater support for edu- cation. Charities, and Tax reform, which covered a variety of govemment programs, policy changes, and educational support that would improve the economic condition of fhe poor. Causes and solutions fhaf appeared only a few times were all categorized into Others.

The entire text of each article and transcript was examined for the attributions of responsibility. Two coders coded the articles and tran- scripts after having conducted a series of fraining and pilot-test sessions. We calculated infercoder reliability by double-coding a random subsam- ple (N = 75 or 15%) of the data. Intercoder reliability corrected for agree- ment by chance (Scott's pi) ranged from .71 fo .89 wifh an average reliabil- ify of .79.

Two coders firsf examined whefher each article and program men- tioned any one or more of the six personal and six societal causes. Each cause was coded as "not present" (0) or "present" (1). Coders then deter-

5 6 8 JOURNALISM & MASS COMMUNICATION QUARTERLY

Causal Responsibility

TABLE 1 Attributions of Causal and Solution Responsibility

(Coding and Intercoder Reliability)

Solution Responsibility

Personal-level causes of poverty (Scott's pi = .76)

Making a bad choice: Poor money management; being lazy; alcohol/drug abuse; criminal record/behavior; etc.

Broken family: Teen pregnancy; being a single parent; divorce; having too many children; etc.

Lack of education: Dropping out of school; un- skilled; etc.

Physical conditions: Disabled; health problems; too old to work; etc.

Immigration: Visa status; language barrier, etc. Others: Being victimized; lack of religious belief;

etc.

Societal-level causes of poverty (Scott's pi = .83)

Bad economy: Recession; unemployment; failed economic policies; etc.

Inadequate pay: Businesses offering unreason- ably low salaries; minimum wage is too low; etc.

Ineffective government aid: Problems in welfare programs; aid programs designed to make people lazy; etc.

Lack of education support: Lack of funding for educating underprivileged children; lack of job education opportunities; etc.

Racism: Discrimination in job opportunity, in compensation; etc.

Others: Problems in the tax system; social Dar- winism; etc.

Personal-level solutions to poverty (Scott's pi = .83)

Making better choices: Planned money manage- ment; stop abusing alcohol/drugs; looking out for job opportunities; etc.

Education: Complete high school; learn job skills; etc.

Others: Abstinence; being reUgious; learn English; etc.

Societal-level solutions to poverty (Scott's pi = .73)

Better economic policies: Create more jobs; boost the economy; better economic policies; etc.

Higher minimum wage: Raise the minimum wage; etc.

Reforming government aid programs: Restructuring aid programs; etc.

Greater support for education: Higher educa- tion opportunities for low-income families; make job education opportunities more available and accessible; etc.

Charities: Religious, charity organizations being responsible for helping the poor; etc.

Tax reform: Fair and reasonable tax rates for low-income families; restructuring the tax system, etc.

Others: Eradicating racism; control illegal immigration; etc.

mined how many menfions of personal and sociefal causes were made in each news reporf. While some articles and programs confained no mention of eifher personal or sociefal causes, fhere were also many items mentioning more than one of fhe six personal and six sociefal causes. In many cases, the same cause was mentioned more than once in a single news sfory. No matter how many mentions were made, we counted fhem as one menfion if they came from the same article or pro- gram. This allowed us fo avoid unnecessarily inflating the number of mentions made of a particular cause. Using the same method, coders also counfed how many personal and sociefal solutions were mentioned in each news story.

TALKING ABOUT POVERTY 569

FIGURE 1 The Amount of Newspaper and TV News Coverage of Poverty

100-

80-

60-

40-

20-

0 -

Jy 1

AKA/ \\' \ ̂—. A

/ \

/ \ \

\

\ .. --̂

i

Mi \ \ X

h \ \ /// y .1 ''""

/ ̂

Newspaper Television

. ^ Newspaper •• Television

19961997 19981999 2000 2001 2002 2003 2004 2005 2006 2007

Year

Note: The amount of newspaper coverage includes articles from the Tulsa Wortd, Birmingham News, Albuquerque journal, Richmond Times Dispatch, the Charleston Gazette, the Santa Fe Neiv Mexican, the Boston Globe, the Denver Post, and Star Tribune matching the key word "poverty" appearing in the headline, after sorting out unrelated items (the first-round filtering). The amount of television news contains news programs from ABC, NBC, CBS, and CNN matching the same keyword.

Findings Figure 1 presents the total number of news stories between 1993 and 2007 matching the keyword "poverty" appearing in the headline. As shown in the figure, the amount of news coverage has been cyclical, with a couple of ups-and-downs during the tifteen-year period. A quick comparison with Figure 2 (poverty rate) demonstrates that the amount of news coverage does not necessarily correspond to changes in the pover- ty rate. While the poverty rate was declining between 1995 and 1999, for example, the amount of news coverage was, in fact, increasing during the same time period. The figures also show that the amount of news cover- age declined between 2000 and 2004, during which the poverty rate was indeed on the rise.

Our research question (RQ) addresses whether certain causes and solutions have appeared more often than others. The first hypothesis (HI), in particular, examines whether the media have focused more on personal causes and solutions than on societal-level attributions. Table 2 shows that both newspapers and television mentioned broken family most

570 JOURNALISM & MASS COMMUNICATION QUARTERLY

FIGURE 2 Poverty Rate*

16.0-

9.0-

"I I I 1 1 1 1 1 1 1 r 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Year

*Source: U.S. Census Bureau (www.census.gov/prod/2006pubs/p60-231.pdf).

often as a personal cause of poverty (sixty out of 506 stories, or 11.9%). Following next were lack of education and physical conditions, which appeared in a total of thirty-three (6.5%) and twenty-five (4.9%) stories respecfively. Making a bad choice and immigration were mentioned less often, both appearing only sixteen (3.2%) fimes. Table 3 shows that the most frequently mentioned societal cause was bad economy {N = 55, 10.9%), followed by inadequate pay (N = 45, 8.9%) and ineffective govern- ment aid programs {N = 44, 8.7%). Lack of education support and racism were mentioned less often (fourteen and nine mentions respecfively), accounfing for only 2.8% and 1.8% of the total news stories.

Table 4 shows that education was the most frequently menfioned personal solution, appearing in forty-one (8.1%) articles and transcripts. Making better choices was found less often (N = 20), accounfing for 4.0% of the entire news stories. When it comes to societal-level attributions (see Table 5), greater support for education and better economic policies were menfioned most often, appearing ninety-four (18.6%) and seventy-three (14.4%) fimes. Attribufions of responsibility to charities or to reforming government aid programs were found less often (fifty-eight and forty-three), accounfing for about 11.5% and 8.5% of the news stories. Higher minimum wage (N = 26) and Tax reform {N = 23) were menfioned far less frequently.

TALKING ABOUT POVERTY 571

TABLE 2 Attributions of Personal-level Gausal Responsibility in News Goverage of Poverty

(1993-2007)

Personal-Level Causes

Media (N) Making Broken Lack Physical Immi- Others Total a Bad Family of Condi- gration

Choice Education tions

Conservative Newspapers The Tulsa World (53) Birmingham News (52) Albuquerque Journal (29) Richmond Times Dispatch (43) Total (177)

Liberal Newspapers The Charleston Gazette (50) The Santa Fe New Mexican (19) The Boston Clobe (52) The Denver Post (46) Star Tribune (51) Total (218) NP Total (395)

Television ABC (15) NBC (16) CBS (34) CNN (46) TV Total (111)

0 3 1 2 6

2 0 0 1 4 7

13

0 0 0 3 3

4 8 4 6

22

7 1 5 4 9

26 48

2 1 5 4

12

5 7 2 2

16

2 1 3 3 3

12 28

0 0 0 5 5

2 5 1 4

12

2 1 0 2 2 7

19

1 2 1 2 6

1 0 3 2 6

1 0 2 1 3 7

13

0 0 2 1 3

0 0 2 4 6

0 0 0 0 0 0 6

3 2 2 5

12

12 23 13 20 68

14 3

10 11 21 59

127

6 5

10 20 41

Media Total (506) 16 60 33 25 16 18 168

Note: Entries are the number of news articles and transcripts mentioning each cause of poverty.

Taken together, references to personal causes and solutions were made a total of 240 times (168 mentions of personal causes and seventy- fwo mentions of personal solufions, see Tables 2 and 4) wifh an average of .47 mentions in each news story (240 mentions in 506 arficles and fran- scripts). Societal attributions of responsibility appeared much more often (202 mentions of societal causes and 375 mentions of societal solutions) with an average of 1.14 mentions per news report (see Tables 3 and 5). A paired-sample t-test indicated that the difference was statistically sig- nificant (t = 10.874, p < .001); but the finding was in fact counter-hypothet- ical. Contrary to what we predicted in HI, the media's attributions of responsibilify were largely sociefal, focusing more on social causes and solutions.

H2a and H2b examine whether conservative and liberal newspapers are different in terms of affributions of responsibilify. Tables 2 and 4 show that conservative papers made sixty-eight mentions of personal causes and twenty-three mentions of personal solutions with an average of .51

572 JOURNALISM & MASS COMMUNICATION QUARTERLY

TABLE 3 Attributions of Societal-level Causal Responsibility in News Coverage of Poverty

(1993-2007)

Societal-Level Causes

Media (N)

Conservative Newspapers The Tulsa World (53) Birmingham News (52) Albuquerque Journal (29) Richmond Times Dispatch (43) Total (177)

Liberal Newspapers The Charleston Gazette (50) The Santa Fe New Mexican (19) The Bosfon Globe (52) The Denver Post (46) Star Tribune (51) Total (218) NP Total (395)

Television ABC (15) NBC (16) CBS (34) CNN (46) TV Total (111)

Bad Econ- omy

6 2 0 2

10

4 2 9 1

14 30 40

3 1 3 8

15

Inade- quate Pay

5 4 2 5

16

2 0 7 1 8

18 34

2 1 3 5 11

Ineffec- tive

Govt. Aid

3 2 2 1 8

4 1

12 3

10 30 38

0 1 3 2 6

Lack of Educa-

tion Supporf

2 4 0 0 6

1 0 2 1 1 5 11

0 0 1 2 3

Racism

0 1 0 1 2

3 1 0 1 0 5 7

0 2 0 0 2

Ofhers

2 1 0 4 7

6 4 4 8 5

27 34

0 1 0 0 1

Total

18 14

4 13 49

20 8

34 15 38

115 164

5 6

10 17 38

Media Total (506) 55 45 44 14 9 35 202

Note: Entries are fhe number of news articles and transcript.s mentioning each cause of poverfy.

menfions per sfory (a fofal of ninefy-one menfions in 177 articles). Liberal papers, on fhe ofher hand, made fiffy-nine menfions of personal causes and fwenfy-nine menfions of personal solufions in fheir 218 arfi- cles, which was an average of .40 menfions in each arficle. Therefore, conservative newspapers menfioned personal causes and solufions slighfly more offen fhan liberal papers (.51 vs. .40). The difference, how- ever, was nof sfatisfically significanf. H2a was nof supporfed. When if comes fo sociefal affribufions (see Tables 3 and 5), liberal newspapers, as predicfed in H2b, made more references fo social causes and solufions (M = 1.39) fhan did conservafive papers (M = .93). This difference was sfatistically significanf (t = 4.080, p < .001), supporting H2b.

Our lasf fwo hypofheses address whefher there is a difference between television and newspapers in the way they attribute responsibility. Television made slightiy more frequent mentions of per- sonal causes and solufions (M = .55) fhan did newspapers (M = .45). The

TALKING ABOUT POVERTY 573

TABLE 4 Attributions of Personal-level Solution Responsibility in News Coverage of Poverty

(1993-2007)

Personal-Level Solutions

Media (N) Making Better Choices

Conservative Newspapers The Tulsa World (53) Birmingham News (52) Albuquerque Journal (29) Richmond Times Dispatch (43) Total (177)

Liberal Newspapers The Charleston Gazette (50) The Santa Fe New Mexican (19) The Boston Globe (52) The Denver Post (46) Star Tribune (51 ) Total(218) NP Total (395)

Television ABC (15) NBC (16) CBS (34) CNN (46) TV Total (111)

2 1 0 4 7

0 0 3 1 5 9

16

0 1 0 3 4

Education

4 5 0 3

12

1 1 5 2 9

18 30

1 1 3 6

11

Others

0 0 2 2 4

0 0 1 1 0 2 6

1 0 1 3 5

Total

6 6 2 9

23

1 1 9 4

14 29 52

2 2 4

12 20

Media Total (506) 20 41 11 72

Nofe.' Entries are the number of news articles and transcripts mentioning each solution to poverty.

difference, however, was not sfafisfically significant. H3a was not sup- porfed. As far as societal attributions, newspapers made more frequenf references to social causes and solutions (M = 1.19) than did television (M = .97), although the size of the difference approached buf did nof reach conventional levels of statistical significance (t = 1.724, p = .085). H3b was supported only marginally.

Discussion Analyzing news articles and franscripfs, we explored how American news media have framed fhe issue of poverfy. More specifical- ly, our study examined how the media presented fhe quesfions of whaf were the causes and how fo fix fhe problem. We also explored fhe notion of frame building, looking af fhe facfors fhat might influence the way an issue is framed. What do our findings fell us about news coverage of poverty? How does this sfudy contribufe fo fhe framing literature?

We first found that fhe amount of poverty coverage was nof close- ly correlated with the acfual prominence of the problem (poverty rate). The amount was rather irvfluenced by newsworthy evenfs fhat could

574 JOURNALISM £r MASS COMMUNICATION QUARTERLY

TABLE 5 Attributions of Societal-level Solution Responsibility in News Coverage of Poverty

(1993-2007)

Societal-Level Solutions

Media (N)

Conservative Newspapers The Tulsa World (53) Birmingham News (52) Albuquerque Journal (29) Richmond Times Dispatch (43) Total (177)

Liberal Newspapers The Charleston Gazette (50) The Santa Fe New Mexican (19) The Boston Globe (52) The Denver Post (46) Star Tribune (51) Total (218) NP Total (395)

Television ABG (15) NBC (16) GBS (34) GNN (46) TV Total (111)

Better Econo-

mic Policies

7 5 1 1

14

11 1 7 6

10 35 49

3 3 3

15 24

Higher Mini- mum Wage

1 0 3 2 6

4 1 4 2 3

14 20

1 0 0 5 6

Reform- ing

Govt.Aid Programs

4 3 5 1

13

8 0 8 4 5

25 38

1 1 0 3 5

Greater ( Support

for Education

7 7

10 4

28

8 3

16 12 11 50 78

1 0 5

10 16

Charities

1

8 9 3 4

24

4 0 7 7 8

26 50

2 1 2 3 8

Tax Others Reform

3 1 2 0 6

4 1 1 0 2 8

14

2 0 0 7 9

10 3 9 3

25

3 4 7 7

10 31 56

0 0 1 1 2

Total

40 28 33 15

116

42 10 50 38 49

189 305

10 5

11 44 70

Media Total (506) 73 26 43 94 58 23 58 375

Note: Entries are the number of news articles and transcripts mentioning each solution to poverty.

draw the public's attention. The large increase in news coverage in 1999 (see Figure 1), for example, can be explained by the fact that the year was when President Clinton began his poverty tour, in which he pro- moted a plan for attracting jobs and investment to the economically depressed regions of the country. The year 2005, which showed anoth- er increase in coverage, was when Hurricane Katrina hit the Gulf Coast, drawing renewed attention to the poverty problem. This finding is con- sistent with the agenda-building literature showing that the amount of news coverage does not necessarily correspond to the actual serious- ness of a problem.*^ Even though poverty has been an issue for a long time, it seems to be the authority of political figures and the newswor- thiness of dramatic events that can define the issue as a social problem important enough to deserve great media attention.

An interesting finding of this study is that the most frequently mentioned causal attribution was broken family, with teen pregnancy or

TALKING ABOUT POVERTY 575

being promiscuous cast as the major cause of poverty. This is not a new finding; previous studies have reported that the media portray the poor not only as lazy, but also as sexually irresponsible."** Frequently linking poverty to irresponsible sexual behavior is problematic for several rea- sons. First, there is little evidence to support a close connection between the two. According to the Alabama Poverty Project, the fertility rate of women living on welfare is lower than the overall fertility rate of American women of child-bearing age."" Second, the claim that links poverty to being sexually irresponsible contains religious overtones, implying that the poor have done something immoral that makes them deserve the poverty in which they live. This unsubstanfiated claim can further reinforce the "welfare queen" stereotype that depicts poor people as mostly single mothers who choose to be in that position. Finally, deriving the cause from immoral behaviors can lead to flawed conclusions. For example, can people avoid becoming poor simply by waiting until after high school to have a child? Interesfingly, while many news stories mentioned teen pregnancy as a cause of poverty, rarely did they discuss abstinence as a way to solve the problem.

Our findings also tell us about the factors that may affect the media's selective uses of certain frames. Given the strong individualism of the American culture, we expected that the media would focus more on individual accountability than on societal responsibilities. Contrary to what we expected, however, mentions of societal causes and solufions significantly outnumbered the references made to personal respon- sibilities. There are several explanations. First, it is possible that the predisposifions of individual journalists play a more important role beyond and above the influence of the individualistic culture. American journalists who cover political news are known to be predominantly liberal in political views when compared to the general population.^" Research also shows that journalists' own political views can influence news coverage, though the influence tends to be subtle rather than overt."''' The liberal ideology may have led journalists to focus on greater government involvement, calling for more societal approaches. Second, poverty stories are often found when the government or community organizations announce new anti-poverty initiatives. Because these initiatives are focused largely on rectifying social and structural sources of poverty, they often provide journalists with a large amount of news material that involves societal causes and solufions. Lastly, a significant number of poverty stories were reported when there was a large-scale natural disaster. Given the nature of the circumstance, it would be unlike- ly for news writers to blame the victims themselves for their situation. Instead, the local government and the community as a whole are often blamed for being unprepared.

Even though the liberal ideology of news writers might be an important factor affecting frame building, it clearly was not the only factor. The overall pohtical leanings of news organizafions (i.e., the view of the publisher) also affected which level of responsibility will be selected. We found that conservative newspapers, when compared to liberal papers, were significantly less likely to menfion societal causes and solutions.

576 JouRNAusM & MASS COMMUNICATION QUARTERLY

That is, the unbalanced emphasis on social responsibility was considerably smaller among conservative newspapers, indicating that organizafional pressures can also play a role. The political orientation of the organization may become a powerful influence as it becomes a form of organizational pressure. It is likely that conservative pub- lishers' deeper financial and organizational resources for skilled man- agement put a limit on how far fo the left their employees can go on the ideological spectrum.™

Episodic framing of the issues, or focusing on personal stories rather than social backgrounds, is a kind of professional routine for the American news media.'' A necessary consequence of episodic framing is to displace attenfion away from larger social conditions, placing people—rather than society—at the center of discussing how to solve a problem. We hypothesized that if fhis routine practice does affect which frames will be selected, television would be less likely than newspapers to emphasize societal causes and solutions. Compared to newspapers, television is much more event-oriented. Television is a story teller. By and large, it does not cover the issues; it tells stories." It is in this process of telling stories that television diverts attention from systemic and institutional responsibilities." Supporting the hypothesis, we found that television was slightly less likely to assign responsibility to society. Though the finding was only marginally significanf, if provides at least some evidence that the professional routine of news media—as being episodic and evenf-oriented—may affect the media's selective use of one frame over another.

While previous research has focused largely on stereotypical portrayals of the poor,'^ our sfudy offers an analysis of how fhe media present the question of responsibility. In particular, we examined an important funcfion of news framing: diagnosing the causes and suggesfing solutions.^^ What then are some potential effects of such framing on the audience? Does news coverage of poverfy affect peo- ple's perceptions of who is responsible? Is this perceived responsibility in turn related to the public's attitudes toward the poor? More important, do the different ways to talk about poverty influence peo- ple's supporf for government aid to the poor? Future research building on our findings needs to explore such consequences of responsibilify framing, looking directly at the impact of news coverage on the public's perceptions and attitudes.

It will also be important to examine whether the typical ways poverty is framed are driven by politicians. To what extent do the media parrot what politicians tell them about poverty? Or is there a mutually reinforcing relationship between politicians' and the media's uses of frames? According to the indexing hypothesis,^* news professionals tend to "index" the range of viewpoints according to the range of views expressed in the mainstream government debate. Other studies, however, reported that news media operate somewhat independently, offen adopfing a sef of frames different from that of politicians and interest groups.^^ Future research needs to explore in detail fhe interplay among the media, politicians, and interest groups

TALKING ABOUF POVERTY 577

in the process of framing building, looking af how these players influence one another to shape the public discourse on the poverty issue.

NOTES

1. U.S. Bureau of the Census, "Income, Poverty, and Health Insurance Coverage in the United States: 2008," prepared by Carmen DeNavas- Wait, Bernadette D. Proctor, and Jessica C. Smith, Bureau of fhe Census, Washingfon, DC, 2009, hftp://www.census.gov/prod/2009pubs/p60- 236.pdf (accessed June 18, 2010).

2. Charles T. Salmon, "Campaign for Social Improvemenf: An Overview of Values, Rationales, and Impacfs," in Information Campaigns: Balancing Social Values and Social Change, ed. Charles T. Salmon (Newbury Park, CA: Sage, 1989), 19-53.

3. Todd Giflin, Inside Prime Time (NY: Panfheon, 1983). 4. Shanfo Iyengar, Is Anyone Responsible?: How Television Trames Politi-

cal Issues (Chicago: Universify of Chicago Press, 1991). 5. Roberf M. Enfman, "Framing of U.S. Coverage of Infernafional

News: Confrasfs in Narrafives of fhe KAL and Iran Air Incidenfs," Journal of Communication 41 (aufumn 1991): 6-27; William A. Gamson and Andre Modigliani, "The Changing Culfure of Affirmafive Acfion," Research in Political Sociology 3 (1987): 137-77; Holli A. Semetko and Patti M. Valkenburg, "Framing European Politics: A Confent Analysis of Press and Television News," Journal of Communication 50 (spring 2000): 93-109.

6. Amy E. Jasperson, Dhavan V. Shah, Mack Watts, Ronald J. Faber, and David P. Fan, "Framing and the Public Agenda: Media Effects of the Importance of the Federal Budget Deficif," Political Communication 15 (April-June 1998): 205-24.

7. Renifa Coleman and Stephen Banning, "Network TV News' Affec- tive Framing of fhe Presidential Candidates: Evidence for a Second-Level Agenda-Setting Effecf fhrough Visual-Framing," Journalism & Mass Com- munication Quarterly 83 (summer 2006): 313-28.

8. Diefram A. Scheufele, "Framing as a Theory of Media Effecfs," Journal of Communication 49 (January 1999): 103-22.

9. Scheufele, "Framing as a Theory." 10. Roger W. Cobb and Charles D. Elder, Participation in American

Politics: The Dynamics of Agenda-Building (Boston: Allyn & Bacon, 1972). 11. Scheufele, "Framing as a Theory." 12. Pamela J. Shoemaker and Sfephen D. Reese, Mediating the Message:

Theories of Infiuences on Mass Media Content (White Plains, NY: Longman, 1991).

13. Scheufele, "Framing as a Theory." 14. Lawrence Wallack, Lori Dorfman, David Jernigan, and Makani

Themba, Media Advocacy and Public Health: Power for Prevention (Newbury Park, CA: Sage, 1993).

15. Sei-Hill Kim and L. Anne Willis, "Talking abouf Obesity: News Framing of Who Is Responsible for Causing and Fixing fhe Problem," Journal of Health Communication 12 (June 2007): 359-76.

5 7 8 JOURNALISM & MASS COMMUNICATION QUASTERLY

16. Herbert Gans, Deciding What's News (NY: Pantheon, 1979); Shoemaker and Reese, Mediating the Message.

17. Sei-Hill Kim, John P. Carvalho, Andrew G. Davis, and Amanda M. Mullins, "The View af the Border: News Framing of the Definition, Causes, and Solutions to Illegal Immigrafion" (paper presenfed af fhe annual meefing of AEJMC, Chicago, 2008).

18. Murray Jacob Edelman, Political Language (NY: Academic Press), 51.

19. Scheufele, "Framing as a Theory." 20. Gamson and Modigliani, "The Changing Culfure." 21. Teun A. Van Dijk, "Structures of News in the Press," in Discourse

and Communication: New Approaches to the Analysis of Mass Media Discourse and Communication, ed. Teun A. Van Dijk (NY: W. de Gruyter, 1985), 69-93.

22. Leon V. Sigal, Reporters and Officials (Lexington, MA: D.C. Heath, 1973).

23. Iyengar, Is Anyone Responsible? 24. Seow Ting Lee and Crispin C. Maslog, "War or Peace Journ-

alism? Asian Newspaper Coverage of Conflicts," Journal of Communica- tion 55 Oune 2005): 311-29.

25. Scheufele, "Framing as a Theory." 26. Robert M. Entman, "Framing: Toward a Clarification of a Frac-

tured Paradigm," Journal of Communication 43 (autumn 1993): 51-58. 27. Sei-Hill Kim, Diefram A. Scheufele, and James Shanahan, "Think

about It This Way: Attribute Agenda-Seffing Function of fhe Press and the Public's Evaluation of a Local Issue," Journalism & Mass Communi- cation Quarterly 79 (spring 2002): 1-25.

28. Shanto Iyengar, "How Citizens Think abouf National Issues: A Matter of Responsibility," American Journal of Political Science 33 (Nov- ember 1989): 878-900.

29. Sfephen Hilgartner and Charles L. Bosk, "The Rise and Fall of Social Problems: A Public Arena Model," American Journal of Sociology 94 Ouly 1988): 53-78.

30. Kim and Willis, "Talking about Obesity." 31. Daniel Wikler, "Personal and Social Responsibilify for Health,"

Ethics & International Affairs 16 (Sepfember 2002): 47-55. 32. William Ryan, Blaming the Victim (NY: Vintage, 1976). 33. Kathryn C. Montgomery, "Promoting Health through

Entertainment Television," in Mass Communication and Public Health: Complexities and Conflicts, ed. Charles Afkin and Lawrence Wallack (Newbury Park, CA: Sage, 1990), 114-28; Lawrence Wallack, "Mass Media and Healfh Promotion: Promise, Problem, and Challenge," in Mass Communication and Public Health: Complexities and Conflicts, ed. Charles Atkin and Lawrence Wallack (Newbury Park, CA: Sage, 1990), 41-51.

34. Mike Budd, Sfeve Craig, and Clay Sfeinman, Consuming Environment: Television and Commercial Culture (New Brunswick, NJ: Rufgers University Press, 1999).

35. Salmon, "Campaign for Social Improvement."

TALKING ABOUT POVERTY 579

36. Iyengar, ¡s Anyone Responsible? 37. Wallack et al.. Media Advocacy. 38. Iyengar, Is Anyone Responsible?; Kim and Willis, "Talking about

Obesity." 39. Linda Jean Kensicki, "No Cure for What Ails Us: The Media-

Constructed Disconnect between Societal Problems and Possible Solutions," Journalism & Mass Communication Quarterly 81 (spring 2004): 53-73.

40. Iyengar, Is Anyone Responsible? 41. Kent Goshorn and Oscar JJ. Gandy, Jr., "Race, Risk and Respon-

sibility: Editorial Constraint in the Framing of Inequality," Journal of Communication 45 (spring 1995): 133-51.

42. The list is available at www.census.gov/hhes/www/income/ income06 / statemhi3.html.

43. This "endorsement-based" categorization of each newspaper might be somewhat artificial, and, as an anonymous reviewer points out, has been driven mostly by politicians and the media. We recognize that our operational definition of being conservative or liberal has a number of limitations and may not reflect reality very well. At the minimum, how- ever, the operationalization may well represent the political view of the publisher.

44. Not all the newspaper articles and television transcripts were made available in LexisNexis for the entire fifteen-year period. Data for some newspapers and television news (Richmond Times Dispatch, the Denver Post, the Charleston Gazette, the Santa Fe New Mexican, ABC, and NBC) were only available from the year 1994 or later.

45. Shoemaker and Reese, Mediating the Message; Wallack et al.. Media Advocacy.

46. Peter Parisi, "A Sort of Compassion: The Washington Post Explains the 'Crisis in Urban America,'" The Howard Journal of Communications 9 Ouly-September 1998): 187-203.

47. Alabama Poverty Project, "Myths and Facts about Poverty and Welfare," Alabama Poverty Project, http://www.alabamapoverty.org/ myths (accessed September 15, 2006).

48. Kenneth Dautrich and Thomas Hartley, How the News Media Fail American Voters: Causes, Consequences, and Remedies (NY: Columbia University Press, 1999); Pew Research Center for People and the Press, "Bottom Line Pressures Now Hurting Coverage, Say Journalists," Pew Research Center, http://people-press.org/reports/display.php3?PageID =829 (accessed September 15, 2006).

49. Marsha Barber and Ann Rauhala, "The Canadian News Directors Study: Demographics and Political Leanings of Television Decision- Makers," Canadian journal of Communication 30 (2, 2005): 281-92; S. Robert Lichter, Stanley Rothman, and Linda S. Lichter, The Media Elite (Bethesda, MD: Adler & Adler, 1986).

50. Robert M. Entman, "Framing Bias: Media in the Distribution of Power," Journal of Communication 57 (March 2007): 163-73.

51. Iyengar, 7s Anyone Responsible? 52. Wallack et al.. Media Advocacy.

580 JOURNALISM & MASS COMMUNICATION QUARTERLY

53. Gitlin, Inside Prime Time. 54. Rosalie A. Clawson and Rakuya Trice, "Poverty as We Know It:

Media Portrayals of the Poor," Public Opinion Quarterly 64 (spring 2000): 53-64; Martin Gilens, "Race and Poverty in America: Public Mispercep- tions and the American News Media," Public Opinion Quarterly 60 (win- ter 1996): 515-41; Parisi, "A Sort of Compassion."

55. Entman, "Framing: Toward a Clarificafion." 56. W. Lance Bennett, "Toward a Theory of Press-State Relafions,"

Journal of Communication 40 Qune 1990): 103-25. 57. E.g., Karen Callaghan and Frauke Schnell, "Assessing the Demo-

cratic Debate: How the News Media Frame Elite Policy Discourse," Political Communication 18 (April-June 2001): 183-212.

TALKING ABOUT POVERTY 581

Copyright of Journalism & Mass Communication Quarterly is the property of Association for Education in

Journalism & Mass Communication and its content may not be copied or emailed to multiple sites or posted to a

listserv without the copyright holder's express written permission. However, users may print, download, or

email articles for individual use.