Abstract
Abstract
Social movements and collective action have often been studied using newspapers for data. Concerns with newspaper data include selection bias, where a subset of accounts, events, or statements are reported, and description bias, which refers to the veracity of accounts and statements. Bias can be created by intentional or accidental misstatements of events and processes, by type of event selected by journalists, and by characteristics of the reporting entity. In this article, bias and scholarly strategies to address bias are discussed in the context of two West Virginian environmental conflicts; one focuses on mountaintop removal mining and one involves the production of the insecticide methyl isocyanate (MIC). Newspaper data is used to explore what activists say motivates their activism toward corporations causing environmental hazards.
Introduction
A
West Virginia was selected as the location for this study because it is a primary source of American coal, it is also a state with infamous conflicts between the working class and industry (coal, chemical production, hydro-power, and paper manufacturing) over quality of life, health, and increasingly, environmental hazards. As a result, studying communities in West Virginia can shed light on labor, public health, and environmental activism. It is in this context that I undertook research on coal production—specifically mountaintop removal mining, and coal's sister industry—chemical production.1 In particular, I am interested in situations where corporations are very important to a low-income area's economy, and some activists are part of local populations that will suffer economically from the corporations' cuts in productions or employment. I asked: What do activists say motivates their activism toward corporations causing environmental hazards? This is not a novel problem but remains difficult to solve methodologically and therefore merits discussion. Below, two environmental controversies in West Virginia are briefly described, after which activist newspaper statements will be discussed, compared to other data and prior work, and critiqued.
What is Happening in West Virginia?
The two West Virginia environmental conflicts examined with newspaper data are tied to mining and to chemical production, respectively. First, activists in and around Whitesville and Naoma, West Virginia are opposed to the Massey Energy, now Alpha Coal Company, practice of mountaintop removal mining to extract coal. Mountaintop removal mining (MTR) is a method used to reach coal deposits by setting off explosives to expose coal seams that are then stripped of coal (West Virginia Coal Association 2009).
Activists began to oppose MTR coal mining because they perceived negative effects on their local culture, economies, health, safety, and ecosystems (Bontrager 2009, Sparki 2010, Simmons Jr. 2005, PR Newswire 2009, Miller 2004, Mountain Justice, Walk 2011, Ward Jr. 2008, Parsons 1999, Shnayerson 2008, Myers 1998, Nyden 2009, Nease 1999, Jarrell 2011, Haltom 2005, Coal River Mountain Watch 2014).
Scientists now argue that MTR causes deforestation, burying and pollution of watersheds, and ecological degradation while the toxins and heavy metals found in coal waste cause damage to kidneys and gallbladders; these toxins and metals are also believed to be carcinogenic and harmful to prenatal health (Scott 2010, Hendryx et al. 2008, Hendryx 2008, Ahern et al. 2011, Hendryx and Zullig 2009, Hendryx 2009, Union of Concerned Scientists 2010). Activists are also joined by researchers who argue that MTR has a deleterious effect on Appalachian economies (Hansen et al. 2008). Activist groups opposed to mountaintop removal mining have formed in Whitesville and Naoma, West Virginia (Coal River Mountain Watch [CRMW] and Mountain Justice [MJ], respectively) to campaign against Massey/Alpha.
The second conflict is in Institute and Dunbar, West Virginia, hereafter referred to as the Kanawha Valley area. In December 1984, one of the world's worst industrial disasters occurred in Bhopal, India, when a Union Carbide chemical plant accidentally released more than 40 tons of methyl isocyanate (MIC), killing 3,800 people and sickening thousands more; it was the worst chemical accident in recorded history (Broughton 2005, Chandrashekharan 2011). The Kanawha Valley hosted the only American facility that produced MIC, making the events in India particularly worrisome to those who lived there (St. John 1985). An activist group formed to protect community health and safety, calling themselves People Concerned About MIC (PCAMIC) (St. John 1985). They waged an ongoing campaign against the production of MIC by several chemical firms. The most recent of these is Bayer CropScience AG; Union Carbide was purchased by Dow Chemical Company on August 4, 1999 and operates units at the Bayer-owned Institute location in Kanawha Valley (Dow Chemical Company 2011). The demographics of the study areas are illustrated in Table 1.
Source: U.S. Census Bureau 2010.
Coal River Mountain Watch, CRMW; Mountain Justice, MJ; People Concerned About Methyl Isocyanate, PCAMIC.
Literature Review
Although there are many cases of newspaper data being useful to researchers, there are a number of methodological issues tied to the use of this type of data. Problems include those pertaining to data collection, in particular selection bias (selectively reporting stories due to competition with other newspapers, reporting norms, and editorial concerns) or description bias—errors in reporting from omissions or missing information, and under-representation of actors, groups of actors, or events (Earl et al. 2004). Newspaper data is influenced by types of events considered to be of interest while also being influenced by political climate and social context—as when an otherwise “newsworthy” story is not covered in favor of bigger or local-to-a-newspaper-headquarters news (Oliver and Maney 2000, Maney and Oliver 2001). Prior work indicates that “newsworthy” in practice, often means “the drama of controversy and polarization” or unusual events (Oliver and Meyer 1999). Additionally, there can be coding and data entry bias on the part of researchers using newspaper data, bias introduced through the use of newspaper indexes, inconsistencies over time in how news is reported, and differences in how local versus national newspapers cover news (Earl et al. 2004). Other concerns include the issue of actors, in this case activists, who speak more than once about the same topic, in or across newspaper articles—if repeat mentions of topics occur, it can distort perceptions of group concern regarding that topic.
Still, prior work suggests that at least for national newspapers, reporting bias tends to be relatively stable over time (Barranco and Wisler 1999) so that the inclusion of national newspaper articles provides “… greater stability and can therefore be taken as one of the best available sources of data on collective protest” (Walker et al. 2008)—it should be noted that the best available data is not necessarily fully representative. Local and regional data have been found to be more variable in reporting over time (Ortiz et al. 2005), but these local and regional sources can provide other value in that national newspapers may overlook smaller scale events. Using varied newspaper data can thus aid in triangulation efforts to increase finding robustness.
Methods
I examined activist quotes in newspapers and interviews to explore what activists said motivated their activism against corporations causing environmental hazards. The following passages will focus on the content analysis of 122 articles in order to observe what 98 activists stated directly about their motivations for social movement participation. The analysis includes 317 statements drawn from newspapers. To supplement the newspaper analysis, selected quotes from the 35 interviews I conducted are used. Statements from activists working with the three primary activist organizations: CRMW (the Whitesville case), PCAMIC (the Kanawha Valley case), and MJ (the Naoma case) are compared to each other and to prior work. The newspaper articles are drawn from state, national, and international newspapers. State papers include the Charleston Gazette, the McClatchy-Tribune Business News, and the Charleston Daily Mail; national papers include the New York Times, the Wall Street Journal, the Washington Post, the Roanoke Times, and the Knoxville News Sentinel; international news coverage was taken from the United Press International.
I use varied newspaper sources, written by various journalists who differ in geographic focus, organization norms, and editorial concerns to limit selection bias via triangulation. Similarly, I hope to lessen descriptive bias via analyzing newspaper articles making multiple accounts of the same events, thereby capturing more details. However, use of multiple sources creates other methodological concerns. In particular, variation in timeframe during which the articles are published, region of publication, and political orientation of a newspaper can all introduce additional sources of bias (Earl et al. 2004). If newspaper articles span a number of years as is true for this study, the temporal variance can be an advantage, since looking at actors' decisions over a period of years makes it less likely that the researcher interprets an atypical statement or action as typical. But temporal variance introduces issues of stability over time—are actors views and actions static over time, would one expect them to be, and how does change over time or lack of it effect study findings? Recognizing that data and the people who analyze it are imperfect, it may be that these methodological concerns cannot be fully resolved. Still, adding additional sources of information from interviews and prior work can partially address these concerns.
To analyze the newspaper statements, I follow Creswell's qualitative research method coding strategy. Creswell (2009) advocates a multi-step analysis process that includes organizing and preparing data via transcription; scanning images and typing up field-notes; reading through all data; coding by grouping words, phrases, or passages according to similarities in content; providing a narrative(s) concerning these codes; and composing an interpretive passage in which the “lessons learned” are summarized. Codes for this study and their definitions are presented in Table 2.
Source: Author, March 20, 2012.
Data Analysis
I discuss the analysis process for this article sequentially and compare the newspaper statements with interview quotes and prior work on newspaper data use in studies of collective action and social movements. Table 3 indicates that for the mountaintop removal mining activists of CRMW and MJ, newspaper statements suggest that they are motivated by attachment to place and also by attachment to nature. Unsurprisingly, the mountaintop removal mining activists are more similar in their views than they are to the anti-chemical production activists, of PCAMIC. Contrastingly, PCAMIC activists primarily make newspaper statements about health and safety concerns, and the need for jobs. But how well do these newspaper statements reflect what activists actually say about their motivation to participate in activism toward corporations causing environmental hazards?
Source: Author, August 13, 2011.
Coal River Mountain Watch, CRMW; Mountain Justice, MJ; People Concerned About Methyl Isocyanate, PCAMIC.
First, to avoid the issue of bias introduced by a newspaper article index, I attempted to capture the entire population of articles regarding the two environmental controversies discussed in this article. I did so by using the database LexisNexis which draws from 15,000 news sources to provide comprehensive news coverage (LexisNexis 2014). Keyword searches to locate articles involved activist groups, locations, and issues for example: “Coal River Mountain Watch,” “CRMW,” “mountaintop removal,” and “Whitesville, WV.” No keyword search will be exhaustive, but looking for newspaper articles using varied keyword and keyword combinations increases the likelihood that target stories will be captured in one of the searches. Using a database that captures different types of articles (local, regional, national, international) can reduce selection bias in that triangulating data can assist in efforts toward representativeness of events (Miles and Huberman 1994).
In terms of how activist statements change over time or the newspaper bias toward bigger more dramatic stories, finer grade analyses and other data such as interview and document quotes can be useful. For example, the newspaper data analysis suggests that CRMW activists are motivated by attachment to place. This finding is supported by prior work, which stated that CRMW started out of a concern over strip-mining of all types—CRMW founder Randy Sprouse worried about the jobs lost when communities switched from underground to strip mining, and about the destruction of homes caused by explosives used in strip mining (Shnayerson 2008). Similar placed-based concerns are in CRMW's mission statement which reads: “The mission of Coal River Mountain Watch is to stop the destruction of our communities and environment by mountaintop removal mining, to improve the quality of life in our area and to help rebuild sustainable communities” (Coal River Mountain Watch 2014).
Yet, there is evidence that the newspaper statements mask variability among activists. Sprouse worried about jobs lost and homes destroyed, while interviewee CRMW activists Debbie Jarrell and Junior Walk explained activists' motivation in the following ways:
I think where they come from…their forefathers were farmers you know, used to living off the land, so it's not so much what they want as far as jobs and stuff, it's what they want to preserve (Jarrell 2011). … they're killing us, and there's no way to get around that. Its murder, mountaintop removal is murder because it poisons our water and poisons our air, I think the cancer, it's…How much more likely is it for someone to get cancer who lives around here? [Asking another staffer] It's about three times more likely. I recently had a scare with it myself.…It wasn't surprising that a 21 year old man could have cancer if he's from here (Walk 2011).
Using the coding definitions in Table 2, the Jarell quote would be defined as an “attachment to place” quote because it speaks about community and community history, while the Walk statement would be coded “health and safety” due to the discussion of cancer and cancer rates. In contrast, newspaper quotes such as the following by CRMW activist Janet Nease express some of the same concerns as her fellow activists, but also highlight discrepancies in the scope of problems that activists are concerned about and on the focus of their attention.
… the devastation that is the legacy of mountaintop removal mining is irreparable. The decapitated mountains can never be restored, and the streams buried under tons of debris can never be replaced.…When the coal companies can no longer mine coal in the most profitable but most destructive manner, they will abandon our state and its people, leaving behind aesthetic, environmental and human devastation. (Nease 1999)
The Nease quote mainly speaks about environmental effects, and so would be coded “attachment to nature” although in the final sentence she touches on community effects i.e., “human devastation” and “abandon our state and its people.”
The newspaper data indicating PCMIC activists are focused on health and safety is supported by other sources, but again, change over time is masked. For example, the recently renamed People Concerned About Chemical Safety's (PCACS) mission statement reads:
People Concerned About Chemical Safety (PCACS) is the corporate successor to, and continuation of, the activities of People Concerned About MIC (PCMIC), an unincorporated organization active in community affairs for over 25 years. PCMIC is a community organization in the Kanawha Valley dedicated to the protection of health and safety of all who reside, work, and study in the vicinity of local chemical plants producing highly toxic chemicals. PCMIC was formed around the time we learned that the same chemical that killed and injured thousands in the Bhopal disaster, methyl isocyanate (MIC), was being produced at our neighborhood plant. (People Concerned About Chemical Safety 2014)
The changing scope of the organization's focus from one chemical, MIC, to general health and safety near chemical plants is likely a reflection of shifting realities: Chemical company Bayer announced that it would stop making MIC altogether by mid-2012 (Ward Jr. 2011) after a 26 year dispute regarding the legitimacy of producing that chemical.
However, common threads are clearly evident in newspaper and interview quotes, and other documents such as the mission statement quoted above—the activists feared chemical effects on health and safety. For example, in a newspaper statement, Pam Nixon, former head of PCAMIC, said: “For decades, people in the Institute area were asking valid health and safety questions, even before the 1984 Bhopal tragedy” (Ward Jr. 2009). I interviewed activist Sue Dixon who explained her motivation to become a PCAMIC activist:
After Bhopal, I became active, you know, all of this…we just kind of flowed together because we realized we were in more danger than they had ever told us…when I was pregnant with my second child, I lived in a house that is two blocks over [from the facility] and I was upstairs, and in the middle of the night I woke up so sick, like sea sickness, that way I know it was from the Sevin plant, and I was so sick and dizzy that I called the plant because I was pregnant. I had never called them before. I called them and asked them what the problem was. I said “My house is filled with these fumes and I am so sick.” And he acted like he didn't know and he promised he would call me back, after I told him I was pregnant, I said, “I'm pregnant and I need to know what to do.” And he said he would call me back.…And that particular child of mine…I thought a lot about it recently, she's 31 right now, but from the time that she was born…she was sickly…her skin wasn't right, there were just different things and she's had ventures into the hospital and the doctors swore they couldn't find out what was wrong with her. So I've often suspected that maybe what she was exposed to, because I've suffered from it, so I…that's just a thought.…So I guess when my children were small and after Bhopal that was when I became more active (Dixon 2011).
Her statement would be coded as a “health and safety” quote due to the discussion of illness for her and her daughter.
Finally, the statements made by CRMW and PCAMIC activists about what motivates their activism is further supported by prior collective action and social movement work which finds that grievances and the political, economic, legal, social, and cognitive environments that would-be activists are embedded within influence the choice to become an activist (Diani 2003, Friedman and McAdam 1992, McAdam 1982, Gamson et al. 1982, Gamson 1997, McAdam and Paulsen 1993, McAdam 2003, McAdam and Snow 1997, McAdam and Snow 2000, Soule 1997, Strang and Soule 1998, Walsh and Warland 1983).
Limitations and Conclusions
In sum, this article suggests that newspaper data can add to an analysis but should be triangulated with other sources of data due to bias from temporal and geographic factors as well as activist within group variance. This finding provides support for the small body of methodological research assessing the value of newspaper data and that finds newspaper data to be useful if imperfect. It also counters work by Ortiz et al. (2005) who suggest that newspaper data is so flawed as to be unusable. The potential and problems of newspaper data, and strategies to address problems in said data, need to be explored in varying contexts, to determine the limitations of this data source; this article contributes to that effort in the context of environmental justice and collective action in Appalachia. CRMW concerns related to attachment to place and PCAMIC concerns related to health and safety are present across time, data source, and individual however nuance, differences in focus, and discrepant views are masked. The key takeaway is that statements in newspapers should be seen as part of a mosaic of inter-related but distinct and shifting pieces of information, rather than as unmitigated fact.
Limitations of this study include the focus on two environmental controversies. Though admittedly exploratory, a study of this kind can be used to generate hypotheses that can be tested in future research and in other cases of environmental or social controversies, and activism. As noted, there are dangers in using activist statements since they may misrepresent themselves, either intentionally or unintentionally and newspaper articles introduce a variety of biases. An article source such as LexisNexis provides substantial newspaper data but may not be completely comprehensive. In this study, an actor may make several statements about the same topic, distorting perceptions of which topics are important to activists. Further, where several activist groups are speaking about their motivation to address the same issue, in this case mountaintop removal mining, there is the possibility that the same activist is involved in multiple activist groups. Finally, the author may have introduced bias through coding and data entry.
I have discussed how different types of newspaper articles bring particular strengths to a study although every type of news article has weaknesses as well. Newspaper data in general is subject to problems of selection and description bias, which can be partially offset by other data sources such as interview and document quotes although these in turn are subject to inaccuracies, limitations, and biases. In cases such as author coding and data entry errors, transparency in data collection and coding can partially allow the reader to identify sources of bias; in the current study, the absence of inter-coder reliability checks is a limitation that needs to be noted as is the possibility that multiple newspaper statements by the same individual on the same topic are used. Using interview data in addition to newspaper data can broaden the range of individuals being quoted, but will skew toward organizers and articulate activists who are willing and able to speak about their motivations. My methodological choice to assign an activist to the first organization that they spoke of being affiliated with temporally potentially introduces yet another source of bias although it provides internal consistency within a study and transparency in regard to research protocol. Finally, sources of bias can be addressed through the questions that researchers ask: here, what activists say motivates their activism toward corporations causing environmental hazards is explored. I do not attempt to deduce what activists are actually thinking which is beyond the scope of what can be learned, using newspapers.
Footnotes
Author Disclosure Statement
The author has no conflicts of interest or financial ties to disclose.
1
Coal is used to manufacture thousands of chemicals and coal-rich areas tend to have other geological raw materials useful in chemical making. Bostic June 14, 2011.
