“No Fracking Way!” Documentary Film,Discursive Opportunity,and Local Opposition against Hydraulic Fracturing in the United States,2010 to 2013

Hydraulic Fracturing and Gasland

Hydraulic fracturing—also known as “fracking”—is a method of extracting natural gas that goes back to the 1940s, although it was in only limited use until recent years. Traditionally, the method involved pumping large amounts of pressurized water—mixed with various chemicals—into porous sandstone or limestone to release natural gas. Starting in the early 2000s, drilling companies developed unconventional methods to harvest natural gas trapped in shale rock. These methods involve drilling horizontal wells that extend from their vertical well shafts along relatively thin, horizontal shale layers. Applying the hydraulic fracturing process to a horizontal well requires much larger volumes of water mixed with sand and chemicals under greater pressure than conventional methods, and this poses engineering challenges and increases the potential for environmental pollution.

Opposition to hydraulic fracturing is motivated primarily by local environmental and health concerns. The environmental risks of fracking, previously off the public agenda as an unknown practice, are now being vigorously debated (Osborn et al. 2011; Wilber 2012). To date, no conclusive research exists to prove that hydraulic fracturing is dangerous. On one side, natural gas industry groups and supporters argue that the risk of water contamination is overblown. For example, the Independent Oil and Gas Association of New York argues that the chemicals used in fracking are safe and are made of “a small amount of dilute, benign additives found in common household products” (Wilber 2012:117). On the other side, environmental activists claim the industry is misleading the public by downplaying the risk of water contamination and other environmental problems, and civil society groups are monitoring water quality for evidence of harms (Jalbert, Kinchy, and Perry 2014). Some grassroots groups, such as Marcellus Protest, claim “there are almost as many violations as there are wells. Numerous incidents of spills, contamination and blow outs have been documented.” ³

Awareness about fracking risks increased rapidly in 2010, when the documentary Gasland—directed by filmmaker Josh Fox—was released. Gasland’s extraordinary potential for raising awareness about the environmental risks associated with hydraulic fracturing comes primarily from a few scenes in which Fox visits residents living near natural gas wells and witnesses them light their (apparently methane-contaminated) tap water on fire. Similar to how pictures of the burning Cuyahoga River energized the 1960s environmental movement, powerful images of burning water from Gasland energized activists. Film critics immediately recognized that Gasland could be used as a mobilizing tool: one critic wrote in early 2010 that “Gasland may become to the dangers of natural gas drilling what Silent Spring was to DDT.” ⁴ Activists also recognized that Gasland had a major impact. As one organizer put it, “Josh Fox’s Gasland blew the doors off what was happening, raising awareness a tremendous amount.” ⁵ In a personal communication, Fox told us, “This [raising awareness] happened all over the place. . . . It would take quite a bit of time to document all of the impacts, but I would say there were dozens of examples of this in specific cases and the cumulative impact was a big factor.”

This documentary’s impact on the movement can be assessed at national and local levels. First, if Gasland had a nationwide effect on the public debate about fracking, then this effect should be strongest when the film was released nationally on HBO and when it was nominated for an Academy Award. ⁶ Second, if Gasland had an effect on local areas, this effect should be strongest in communities where the movie was screened, and soon after it was screened. ⁷ Given that multiple issues compete for public attention at any given moment, it is extremely difficult to maintain high levels of awareness about new issues. Any discernible effects of films and other artifacts on activism are thus likely to last for a relatively short period.

Discursive Opportunities: Toward a Richer Understanding of Public Discourse

Most research on framing processes examines how collective action frames contribute to recruitment and collective identities (for a review, see Benford and Snow 2000), but a number of recent studies have generated critical insights into how frames can be linked directly to social movement outcomes (Bail 2012; Cress and Snow 2000; McCammon et al. 2008; McCammon et al. 2007). In so doing, these studies draw attention to the ways contentious collective actors are subject to features of the broader public discourse (Steinberg 1999). These interests have congealed into a focus on how social movement frames, to have policy effectiveness, must not only resonate with potential activists but must also connect with broader themes in public discourse. When a movement’s ideas are congruent with these broader public discourses, actors benefit from “discursive opportunities,” or moments when movement ideas align with what the culture at large deems to be sensible, realistic, or legitimate (Koopmans and Statham 1999; McCammon et al. 2007).

A limitation of these studies, however, is that their conceptualization and measurement of public discourse often privileges discourses represented in the mass media without sufficiently investigating mass discourse more broadly. Understanding the differences between coverage of critical social issues in the mass media versus social media offers insight into a number of key questions in the sociology of the media (for reviews, see Schudson 1989, 2002). Most directly, methods of news-gathering and dissemination are seen by media sociologists as critical to the content of coverage on a topic (e.g., Gans 1979; Gitlin 1980; Tuchman 1978). The bureaucratic routines of traditional news-gathering, for instance, encourage a reliance on official sources and a general deference to official statements (Fishman 1980). When covering contentious politics or protest events, this should mean that events involving social movement organizations (SMOs) will gain greater coverage than those that do not (Earl and Kimport 2011); also, the disruptiveness, resource mobilization, and policy environment surrounding protest groups shape prominent newspaper coverage (Amenta et al. 2009). When covering events associated with controversial or environmentally risky projects, mass media act as either social amplification or social attenuation stations (Flynn, Slovic, and Kunreuther 2001; Pidgeon et al. 2003). It is also well known that the predominant norm of conventional news reporting is to seek out a balanced perspective in reporting (Benson 2004; Gamson and Modigliani 1989), even if this means including a certain amount of false equivalence.

Contemporary social media sources like Twitter and Facebook do not share these features, partly because the work of reporting on events of the day is distributed, non-bureaucratic, and mainly done by lay participants (Chadwick 2013; Kreiss et al. 2011). Furthermore, activist groups often communicate directly through these services, such that SMOs become less important as spokespeople (Bennett 2003); this leads to what some call a “personalization” of collective action (Bennett and Segerberg 2013). Twitter, for instance, is now widely recognized as an effective mechanism on its own for generating and amplifying claims-making among collective actors (Bennett and Segerberg 2012; Hussain and Howard 2013; Tufekci and Wilson 2012). Because social media provides the capacity to broadcast ideas and to organize directly through online means (Shirky 2008), coverage of the risks posed by technological practices such as hydraulic fracturing may be more substantial and more negative in social media than in mass media.

Another important omission of media and social movement studies is an analysis of changing discursive opportunities. We explore the complex nature of movements’ discursive opportunities ranging from general to specific: (1) initial online public attention to an issue; (2) evolution of social media chatter about, and mass media coverage of, a topic; (3) identification of problems associated with a topic in social and mass media (analogous to diagnostic framing); and (4) discussion in social and mass media of what can be done to remedy the issue (analogous to prognostic framing). ⁸

Related to the argument above, existing research on mass media and discursive opportunities does not consider in sufficient depth how movement-related artifacts serve as key sources of changing discursive opportunities. ⁹ This is an important omission because, as research on political opportunity structures shows, social movements can sometimes create opportunities for themselves and for other movements at later points in time. Indeed, as Kriesi (2004:79) notes, “episodes of contentious interaction are likely to modify the relevant configuration of actors and, thus, to change the specific opportunities for future options for collective action.”

We argue that social movements use cultural artifacts to influence public debates and capitalize on discursive opportunities in multiple public spheres. Specifically, we theorize that discursive opportunities are made possible by the national release of documentaries and their nomination for major awards. In this case, Gasland is thought to have had a large impact on building the movement because it identified water pollution as a serious problem and called for organizing, helping to counter dominant frames about the safety of natural gas drilling.

Cultural Products, Mobilization, and Political Outcomes

Analysts of contentious politics argue that movements use cultural artifacts to advance their cause and contribute to positive social change. Movement participants introduce new ideas through such products as films (Andits 2013; Whiteman 2003), music (Danaher 2010; Eyerman and Jamison 1998; Roscigno and Danaher 2001; Roy 2010), and books (Meyer and Rohlinger 2012). Existing research, however, focuses primarily on social movements’ impact on the broader cultural context (Berezin 1994; Eyerman and Jamison 1995; Hanson 2008; Lipsitz 2000; McAdam 1994; Steinberg 2004; Zolberg 1997). For example, studies show that abolitionists introduced African American spirituals into white mainstream culture (Cruz 1999); leftist artists had a major influence on American painting, sculpture, literature, and theater (Hemingway 2002); and the women’s movement of the 1960s and 1970s shaped American poetry (Reed 2005). What is frequently underappreciated in these accounts is an analysis of how social movements “do culture” and its consequences. This is consequential, because movements that bring people together through poetry reading or singing can solidify commitment and mobilize actors toward achieving goals (Roy 2010).

Looking at films in particular, social scientists suggest that activist cinema can have substantial effects, but there is only limited empirical evidence for these arguments to date. For example, Whiteman (2003) argues that the 1992 documentary From the Ground Up, which examined the environmental impacts of mining in Wisconsin, fostered local opposition against mining in Wisconsin through video screenings in diverse venues: bars, county fairs, churches, and sporting group meetings. Leiserowitz (2004) uses survey data to show that individuals who watched The Day After Tomorrow, which depicts fictional disasters spawned by climate change, were more likely to perceive global warming as a threat and more willing to take action. Jacobsen (2011) conducts what is arguably the most rigorous analysis of activist cinema, exploiting spatial variation in the release of former vice-president Al Gore’s climate change documentary An Inconvenient Truth, and shows that the documentary caused individuals to purchase carbon offsets. A variety of other accounts show that films and documentary programming may influence social processes. ¹⁰

Public screenings of documentaries are a more moderate protest form, consistent with other information-distribution strategies such as teach-ins, tabling, leafleting, and holding press conferences to educate the public. ¹¹ But the moderate nature of screenings need not imply they will have little impact. Documentaries can be used to overcome quiescence, or the absence of grievance perception in the face of significant inequality. The fact that individual grievances about environmental problems may exist in a community where polluting industries are located does not mean grievances will be expressed collectively (Crenson 1971; Gaventa 1982; Roscigno 2011). Numerous environmental justice studies show that residents “endure years of uncertainty in the face of contradictory evidence on the health and environmental risks posed by contaminants” (Aronoff and Gunter 1994:243). Because scientists and officials often cannot provide definitive answers about risks, ambiguity and conflicting interpretations frequently emerge (Auyero and Swistun 2008; Brown, Kroll-Smith, and Gunter 2000; Brown and Mikkelsen 1990; Elliott and Frickel 2013; Fowlkes and Miller 1982; Kroll-Smith and Couch 1993; Kroll-Smith, Couch, and Levine 2002). Even when information is made available, it may be presented using scientific jargon that is not accessible to many citizens (Freudenburg and Jones 1991). Moreover, in an unequal political system, citizens may distrust information presented by experts (Cable 2012). Documentaries created by non-experts may present facts in a more accessible manner and overcome such problems of mistrust.

In the case of Gasland, filmmaker Josh Fox attempted to alter public perceptions of social reality and victimhood by showing evidence that fracking can contaminate drinking water, proximity to hydraulic fracturing wells puts public health at risk, and the natural gas industry takes advantage of landowners who sign leases—which generally come with non-disclosure agreements (NDAs), a topic we will return to—without fully understanding the consequences. Gasland also legitimizes anti-fracking activism, prescribes courses of action, and calls for mobilizations: the filmmaker argues that local opposition is not only legitimate but also necessary, given that natural gas drilling is exempt from the Safe Drinking Water Act and government regulators appear unwilling to protect affected communities. Additionally, Gasland has sustained the movement against fracking through screenings in numerous communities across the country. Indeed, the documentary was instrumental in connecting individuals interested in joining the opposition against fracking. The film’s website has links that allow people to share information through Twitter, Facebook, and YouTube, as well as links to contact elected officials and participate in local events. The website tracks participation in actions to “help prevent gas drilling from harming your community.” By the end of November 2013, more than 183,900 actions were reportedly taken through this website.

Like the proverbial tree that falls in the forest, Gasland screenings may have no effect because activists are not present. We argue that this is unlikely: activists are probably involved in organizing local screenings, because the documentary was not shown in large commercial cinemas. Moreover, even if anti-fracking activists did not organize the screenings, they likely knew about the screenings and promoted them in their local communities. Therefore, we argue that Gasland screenings influenced local mobilizations against fracking, but we expect the effect of these screenings was relatively short term, that is, it lasted only a few weeks or months. At any given moment, multiple issues compete for public attention (Baumgartner and Jones 1993), which makes it extremely difficult to maintain high levels of public interest in new issues and to organize long-term collective actions. Indeed, studies of local opposition against energy projects find that not all at-risk communities experience mobilizations and, when they do occur, local mobilizations are usually short lived (McAdam and Boudet 2012).

We also expect that screenings influenced the adoption of anti-fracking policies in a number of ways. First, screenings likely affect the adoption of local bans because they increase the number of mobilizations. Communities with more mobilizations will likely have a higher level of commitment to the cause, even if the number of participants in these events is not very large. Second, screenings of Gasland are likely to attract new activists and sympathizers. According to Josh Fox, screenings frequently involved hundreds of people, and many viewers likely became sympathetic to the cause after seeing the documentary. Indeed, after screenings many people expressed outrage and their determination to “spread the word,” as evidenced in statements such as: “I had no idea,” “I can’t believe they’re getting away with this,” and “I’m going to tell everyone I know to see this film.” ¹² Third, communities with screenings likely have a local political context that is favorable to activism due to the presence of influential allies. For example, councilmembers from cities such as New York and Pittsburgh organized screenings. ¹³ The effect of screenings on local policies can be direct or indirect.

Research Design

We use two separate sets of analyses. First, we explore how a documentary represents the opening of discursive opportunities nationwide. Several theoretical and methodological concerns motivate our choice of examining discourses in what we conceptualize as distinct public spheres. Like previous scholars, we are attentive to the public discourses reported in the media and their particular influences on the political impacts of movement action. Indeed, a primary way that movements “enter the public sphere” is through appearing in media reports (Oliver and Myers 1999). Yet, movements are also enmeshed in broader public discourses that extend well beyond traditional media. Movements may find that discursive opportunities also exist when cultural products on themes related to the movement become popular online. Indeed, recent research shows that online public attention to a social movement—as measured by Internet searches for terms associated with the movement—is a useful predictor of movement emergence (Vasi and Suh 2012). Additionally, movements may find that discursive opportunities exist when relevant cultural products become important topics of social media discussion. Of particular importance is the Twittersphere, or the postings made on the social media website Twitter, which provides a space where people often share opinions and disseminate political information (Gerbaudo 2012; Howard and Hussain 2011; Papacharissi and Oliveira 2012; Vasi and Suh 2012). Moreover, unlike Facebook and other social media platforms, Twitter facilitates the downloading of large volumes of messages on a certain topic.

Next, we examine how the documentary may have influenced local mobilizations and policymaking. We focus on cities within the Marcellus Shale states (Pennsylvania, Ohio, New York, and West Virginia). The Marcellus Shale is the largest shale formation in the United States (Kargbo, Wilhelm, and Campbell 2010), it currently generates more than twice as much natural gas as the next largest formation (Eagle Ford in Texas), and the area was the primary focus of Gasland. Despite its incredible potential for natural gas production, fracking in the Marcellus is a relatively new phenomenon in the past decade. This formation has also had a higher growth rate than all other formations, increasing nearly 13-fold between January 2007 and March 2015. ¹⁴ Drilling in the Marcellus Shale has grown rapidly, and this activity has apparently served as a “suddenly imposed grievance” (Walsh 1981) for residents. Given the vast gas reserves in this shale play, the prospect of long-term societal changes due to fracking is a significant component of public discourse surrounding drilling in the Marcellus.

Also, in models reported in the online supplement for this article, we took additional steps to account for the endogenous selection of Gasland screenings, examining differences between matched cities with and without screenings.

In each of our models of local mobilizations and fracking bans we focus on the city level because the processes we wish to explain occurred primarily within cities; it is reasonable to assume that activists’ city-level organizational activities were more closely associated with a city’s politics than with a state’s politics. To date, only New York and Vermont have adopted a state-level moratorium or freeze on hydraulic fracturing; more than 240 cities have adopted local bans in the Marcellus Shale region. Our dataset includes all cities in the Marcellus Shale states—a total of 3,322 cities—observed between July 29, 2010, and May 31, 2013; a total of 1,038 city-day observations. ¹⁵ The most fine-grained measures were available at the day level; city-days are thus the appropriate units of observation. The rich history of local activism and the decentralized nature of politics in the United States present us with another motivation for choosing the city-level.

Independent Variables

Regarding the documentary, we utilize data about local Gasland screenings. We obtained data from Josh Fox about all screenings of the documentary in the United States. These screenings were organized by individuals and nonprofits in local libraries, cinemas, schools, and other public spaces. In our models, we examine the effects of Gasland screenings using measures of the effects of screenings in the short to medium term (i.e., from a few weeks to a few months post-screening) and within a close radius (five miles between the city where a screening took place and other cities). ²⁶ Some screenings were initiated by Fox and others affiliated with the film; other screenings were initiated by local community members who sought a copy to screen at a local theater or other public venue (Josh Fox, personal communication, March 17, 2014). In a personal communication, Fox said he aimed to inspire local activists to initiate local fracking bans and to give communities the tools they would need to go forward. These local screenings were carried out on a variety of dates starting in August 2010, thus making it difficult to see the effect of these screenings on Google searches or Twitter chatter. Indeed, it is unlikely that the effect of local screenings can be detected in the Google or Twitter data (which is aggregated at the national level) because, although the HBO release has reached millions of viewers, each screening has reached at most a few hundred viewers. ²⁷

We also include a measure of the overall volume of fracking-related Twitter chatter in a given city-day. We coded all Twitter users according to the location they entered in the “location” field associated with a user’s Twitter handle at the time each respective tweet was posted. We first excluded all tweets not issued from the eastern time zone, because all four Marcellus states are in that location; among this subset, 87 percent of tweets were associated with a user with an entry in the location field. We then coded these entries by city and state, finding that approximately 40 percent of tweets with an entry in the location field were posted by a user account located in New York, Pennsylvania, Ohio, or West Virginia. We then converted these tweets into a city-day sum total of fracking-related tweets per 10,000 people, logged the measure, and lagged it by one day. ²⁸

We control for the adoption of municipal ordinances and population size, because larger cities are more likely to have a critical mass of citizens who mobilize against fracking. We measure population size using data from the 2010 U.S. Census, and we apply the natural logarithm to correct for skewness. We control for the dominant political ideology, because environmental campaigns are mainly associated with left-of-center politics (Brulle 2000). We measure left-of-center political ideology using the percentage of votes for Barack Obama, at the county level, during the 2008 presidential election. We control for unemployment, because cities with high levels of unemployment are less likely to oppose drilling for natural gas. Indeed, the natural gas industry frequently touts hydraulic fracturing as beneficial for local economies. We also control for income, because cities with high-income residents are more likely to organize Not In My Back Yard (NIMBY) campaigns (Bullard 2000). We measure unemployment rate and income using data from the U.S. Census Bureau about the unemployment rate and median income in 2010.

We control for the presence of the oil and gas industry using data from the Bureau of Labor Statistics about the total number of people employed in the oil and gas extraction industry in a year, by county. We standardized this measure per 10,000 and lagged it by one year to avoid simultaneity bias. We include as a control whether a city is located directly above the Marcellus. We expect cities located directly above the Marcellus Shale will perceive a greater risk from drilling than cities located elsewhere. ²⁹ Data about Marcellus Shale locations come from maps developed by state agencies. ³⁰

We include measures of proximity to natural gas wells because this might (1) increase citizens’ grievances and lead to mobilizations and the adoption of bans, or (2) depress activism because individuals who live very close to wells have likely leased their land and signed the associated non-disclosure agreements (NDAs) that prevent their activism, or they benefit financially from the industry. We obtained data on permits from each state’s department of natural resources/environmental protection. ³¹ Using these data, we calculated the sum of active hydraulic fracturing wells within a specified radius from each city in the previous year. We experimented with various radii, but for simplicity we include only a five-mile radius. ³² To calculate the sum of all wells we used GeoNear, a Stata module that allows the calculation of distance-based variables; these are log-transformed.

We also include a measure of water contamination events due to fracking, and here we use the same five-mile radius as we do for proximity to wells. To create this measure, we relied on state disclosures on fracking regulatory violations from Pennsylvania and Ohio, combined with an MIT research team’s primary systematic report on all fracking violations to date (Moniz et al. 2011) and national reports on fracking-related water contamination from Earthjustice. ³³ We control for the state in question because each state has a specific regulatory and economic context. We include dummy variables for New York, Ohio, and Pennsylvania (West Virginia is the reference category). We include a measure of environmental organizations, because communities with a strong environmental movement infrastructure are likely aware of hydraulic fracturing’s environmental impacts and possess resources for mobilizing against this practice. ³⁴ We also control for the density of overall nonprofit organizations as a measure of general civic and political capacity of communities, consistent with the finding in many studies that high general civic capacity represents a powerful force to counterbalance industry interests in the siting of environmental bads (Aldrich 2008; McAdam and Boudet 2012; McAdam et al. 2010). All data come from the National Center for Charitable Statistics (NCCS): the environmental groups include all nonprofits categorized in NTEE section C (environmental), and the general civic measure includes all nonprofits in the NCCS overall. ³⁵

Finally, we generated a diffusion variable to examine whether previous adoptions of bans in geographically proximate cities influence the passage of local bans. We experimented with different radii for proximity (e.g., 30, 60, or 100 miles) as well as different functions (cumulative, an inverse square root of distance, and various decay functions) that model how the influence of proximate cities decreases with geographic distance. ³⁶ Alternative specifications produced similar results, so we describe results from the inverse square root of distance measure with a 100-mile radius, because this functional form was used in previous research on diffusion (Andrews and Biggs 2006; Hedström, Sandell, and Stern 2000). Table 1 shows the means, standard deviations, and correlations for all variables.

OPEN IN VIEWER

Table 1.

Means, Standard Deviations, and Correlations

	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17
1. Gasland same day (5m)	.00	.01	1
2. Gasland 2 months (5m)	.01	.09	.02	1
3. Gasland 4 months (10m)	.01	.14	.02	.72	1
4. Population (ln)	7.54	1.43	.01	.08	.10	1
5. Votes for Obama	.48	.08	.01	.07	.09	.36	1
6. Unemployment	8.86	1.97	−.01	−.03	−.04	−.19	−.26	1
7. Income (ln)	168.0	74.14	.01	.04	.05	.30	.26	−.47	1
8. Oil and gas industry	.47	.49	−.00	−.00	−.00	−.16	−.05	−.12	−.28	1
9. Marcellus location	2.16	11.14	−.00	−.00	−.01	−.08	−.14	−.04	−.18	.11	1
10. Location: NY	.29	.45	.00	.02	.03	.17	.30	−.31	.39	−.15	−.10	1
11. Location: OH	.28	.45	−.00	−.02	−.03	−.01	−.11	.66	−.15	−.37	−.06	−.40	1
12. Location: PA	.34	.47	.00	.01	.01	−.09	−.06	−.27	−.01	.49	.04	−.46	−.46	1
13. NGOs (all)	1.78	3.08	−.00	−.01	−.02	−.11	−.04	−.05	−.01	.03	.00	.09	−.08	−.01	1
14. Environmental NGOs	.05	.34	−.00	−.01	−.01	−.06	.01	−.03	−.00	.03	−.01	.05	−.05	.01	.35	1
15. Proximity to wells (5m)	106.6	454.2	−.00	−.01	−.02	−.10	−.10	−.07	−.12	.24	.30	−.14	−.14	.31	−.02	−.01	1
16. Proximity to water c. (5m)	.10	.61	−.00	−.01	−.01	−.03	−.01	.04	−.10	.06	−.00	−.10	.07	.04	.07	.14	−.01	1
17. Twitter messages	4.51	83.34	.00	.00	.00	.00	.01	−.01	.00	.02	.00	.03	−.02	−.01	.05	.09	.01	−.00	1

Results

Changing Discursive Opportunities

We begin by exploring trends in public attention toward fracking. Figure 1 displays trends in overall public attention toward fracking as measured via Google web searches only for “fracking,” and public attention toward fracking as measured via Google web searches only for “Gasland.” As predicted, we find that online public attention toward fracking—as measured by Internet searches—increases immediately after Gasland’s release and award nominations. Gasland contributes to an increase in public attention toward fracking in two ways: first, Internet searches for “Gasland” increase when the documentary is released and nominated—we call this the Gasland-specific public attention to hydraulic fracturing; second, Internet searches for “fracking” increase after the documentary is released and nominated for an Academy Award—we refer to this as fracking-specific public attention.

OPEN IN VIEWER

Figure 1.

Gasland’s Influence on Online Public Attention

These two contributions are apparent in Figure 1. First, searches for the documentary peak in June 2010, when it is released on HBO, and in February 2011, when it is nominated for an Oscar award. Remarkably, Internet searches for the documentary surpass searches for fracking in June and July 2010. For most months, searches for Gasland account for less than 5 percent of the highest value of overall public attention to fracking, but in June 2010 they account for over 40 percent of the searches, in July 2010 they account for approximately 24 percent, and in February 2011 they account for about 20 percent. Second, Internet searches for fracking increase immediately after the HBO release and the Oscar nomination. When examining searches for both fracking and Gasland, we see a clear increase in searches for these terms after the release and nomination.

Figure 1 also displays trends in overall online public attention toward fracking as measured by YouTube searches for “fracking” and for “Gasland.” Again, we find that overall public attention toward fracking—as measured by YouTube searches—increases over time, particularly after Gasland’s release and award nominations. Gasland contributes to an increase in public attention toward fracking in two ways. First, YouTube searches for “Gasland” increase when the documentary is released and nominated—the Gasland-specific online public attention to hydraulic fracturing. Figure 1 shows that YouTube searches for the documentary peak in June 2010, when it is released on HBO, and in February 2011, when it is nominated for an Oscar award. Interestingly, YouTube searches for the documentary surpass overall searches for fracking not only during these two critical months but throughout the entire period from January 2010 until May 2011. This is most likely because YouTube is a video-sharing website; therefore, searches for a documentary are likely to be higher on YouTube than on the web. Second, YouTube searches for “fracking” increase after the documentary is released and nominated—the fracking-specific public attention to hydraulic fracturing. Indeed, the findings presented in Figure 1 are consistent with experiences described in the field. As Gasland creator Josh Fox told us in a personal communication, “When Gasland hits, it changes the game on the word ‘fracking,’ which never leaves the media after that point.”

Next, we explore trends in social media chatter about fracking. We find that chatter in the Twittersphere about fracking increases over time, particularly after Gasland is released on HBO and is nominated for an Oscar award. Figure 2 displays trends in social media chatter about fracking as measured by Twitter messages about fracking or about Gasland. The figure shows that the percentage of Twitter messages about Gasland is larger than the percentage of Twitter messages about fracking when Gasland was released, in June 2010. While the percentage of Twitter messages about Gasland is not larger than that of Twitter messages about fracking in February 2011, when Gasland is nominated for an Oscar, it is relatively close. The release and nomination events are accompanied by an increase in chatter not only about Gasland but also about fracking: chatter increases by approximately 6 percent after the release and by approximately 9 percent after the nomination (compared to the previous month).

OPEN IN VIEWER

Figure 2.

Gasland’s Influence on Twitter Chatter about Fracking

We now examine trends in mass media coverage of fracking. Figure 3 shows the number of newspaper articles that mention “fracking” or “Gasland” between 2010 and early 2013. As predicted, we find that mass media coverage of fracking—as measured by newspaper articles—increases over time, particularly after Gasland’s release and award nominations. Gasland contributes to an increase in coverage of fracking in two ways: first, coverage of Gasland increases when the documentary is released and nominated—we call this the Gasland-specific coverage of hydraulic fracturing. Remarkably, more than half of the articles that mention fracking in June 2010 and January 2011 also mention Gasland. Second, coverage of fracking increases after the documentary is released and nominated for an Academy Award—we refer to this as the fracking-specific coverage.

OPEN IN VIEWER

Figure 3.

Gasland’s Influence on Mass Media Coverage of Fracking

Next, Figure 4 examines changes in social and mass media discussion about fracking—specifically, how diagnostic and prognostic frames emerged through a focus on water problems and on bans and moratoria immediately after Gasland’s release and award nominations. Our automated content analysis reveals that Twitter mentions of the word “water” increase after Gasland is released on HBO (from 1.2 percent in May 2010 to 1.6 percent in June 2010) and after it is nominated for an Oscar (from 1.5 percent in January 2011 to 1.8 percent in February 2011). At the same time, newspaper mentions of the word “water” decrease after Gasland is publicly released (from 2 percent in May 2010 to 1.1 percent in June 2010) and after it is nominated for an Academy Award (from 1.3 percent in January 2011 to 1.1 percent in February 2011). Twitter mentions of the words “ban” and “moratorium” increase after Gasland is nominated for an Oscar (from .3 percent in January 2011 to .6 percent in February 2011) but do not change after Gasland is released on HBO (.3 percent in May and June 2010). Newspaper mentions of these words increase very little after the documentary is released (from 0 to .1 percent) but decrease after it is nominated for an academy award (from .2 to .1 percent). Figure 4 shows that, overall, these words are used less frequently in mass media than in social media. Twitter mentions of bans and moratoria are highly variable over time, reflecting concentrated periods of activity when particular fracking bans were being proposed, debated, and in some cases enacted.

OPEN IN VIEWER

Figure 4.

Gasland’s Influence on Social and Mass Media Discussion about Fracking over Time

Local Mobilizations and Policymaking

Table 2 displays results of models that estimate the effects of Gasland and covariates on mobilizations. Table 3 displays results of a set of models that estimate effects of the anti-fracking events (and covariates) on the passage of local moratoria against fracking. The models in Table 2 are from conditional gap time event history analysis models (clustered on city and stratified by event number); in Table 3, results are from single-event event history models. In both tables, models include measures of proximity to permitted fracking wells and proximity to water contamination accidents (within a five-mile radius). ³⁸

OPEN IN VIEWER

Table 2.

Effects of Gasland Screenings on Anti-fracking Mobilizations in the Marcellus Region States

	Model 1	Model 2	Model 3
Gasland Screenings (five miles)
Four months after, not including day of screening			.171 * (.075)
Two months after, not including day of screening		.375 *** (.099)
Same day of screening	2.669 *** (.806)
Controls
Twitter chatter, total number of messages	.052 *** (.010)	.053 *** (.010)	.053 *** (.010)
Population (ln)	.940 *** (.167)	.935 *** (.163)	.938 *** (.167)
Votes for Obama	−2.230 * (.974)	−2.304 * (.983)	−2.293 * (.963)
Unemployment	−.228(.123)	−.236(.126)	−.234(.123)
Income (ln)	−.001(.001)	−.001(.001)	−.001(.001)
Marcellus location	2.083 *** (.379)	2.119 *** (.392)	2.111 *** (.378)
Oil and gas industry	.011 * (.005)	.011 * (.004)	.011 * (.005)
New York	.569(.511)	.563(.514)	.570(.513)
Ohio	1.302 * (.628)	1.354 * (.641)	1.343 * (.624)
Pennsylvania	1.717 *** (.529)	1.727 *** (.531)	1.728 *** (.528)
NGOs (all)	.154 *** (.020)	.153 *** (.019)	.154 *** (.020)
Environmental NGOs	−.111(.134)	−.111(.134)	−.112(.134)
Proximity to wells (five miles)	−.003 ** (.001)	−.003 ** (.001)	−.003 ** (.001)
Proximity to water contamination (five miles)	.018(.043)	.017(.043)	.017(.043)
N ObservationsChi Square	3,448,236514 ***	3,448,236160 ***	3,448,236113 ***

Note: Conditional gap time models clustered on city and stratified by event number.

*

p < .05; ^**p < .01; ^***p < .001 (two-tailed tests).

OPEN IN VIEWER

Table 3.

Effect of Gasland Screenings and Anti-fracking Mobilizations on Anti-fracking Municipal Bans in the Marcellus Region

	Model 1	Model 2	Model 3
Mobilizations and Screenings
Anti-fracking events (ln)			.765 *** (.242)
Gasland screening five miles (ln)		.634 * (.266)	.511(.262)
Controls
Proximity to previous adopters (100m; inverse sqr. dist.)	.266 *** (.048)	.257 *** (.048)	.266 *** (.049)
Twitter chatter, total number of messages (ln)	.035(.034)	.036(.034)	.024(.036)
Population (ln)	.577 *** (.077)	.538 *** (.078)	.449 *** (.083)
Votes for Obama	−.533(1.428)	−1.533(1.482)	−1.332(1.480)
Unemployment	−.246 * (.115)	−.251 * (.113)	−.240 * (.112)
Income (ln)	1.448 *** (.262)	1.457 *** (.265)	1.398 *** (.266)
Marcellus location	−.001(.002)	−.001(.002)	−.001(.002)
Oil and gas industry	−.042(.041)	−.044(.043)	−.044(.042)
New York	1.408(.841)	1.436(.842)	1.256(.844)
Ohio	1.620(.885)	1.720(.882)	1.617(.881)
Pennsylvania	−.445(.862)	−.487(.862)	−.944(.887)
NGOs (all)	.101 *** (.020)	.101 *** (.020)	.093 *** (.022)
Environmental NGOs	.005(.229)	.001(.224)	.008(.238)
Proximity to wells (five miles)	−.002(.002)	−.002(.002)	−.002(.002)
Proximity to water contamination (five miles)	.350(.259)	.351(.251)	.358(.264)
N ObservationsChi Square	3,401,471220 ***	3,401,471225 ***	3,401,471233 ***

Note: Single-event Cox models.

*

p < .05; ^**p < .01; ^***p < .001 (two-tailed tests).

Results in Table 2 show that screenings of Gasland in a given community lead to a significant increase in anti-fracking events in the relatively short term, and this is true regardless of how the model specifies proximity to active wells or to wells that have contaminated water. Results in Model 1 show that local screenings are associated with local mobilizations in the same day. Results in Models 2 and 3 show that the effect of Gasland does not last very long if we are not including the day of the screening: for example, the effect is significant (p < .001) when it is assumed to last two months, and it is weaker but still significant (p < .05) when it is assumed to last four months after the screening. Thus, the effect of Gasland screenings does not last much beyond four months. ³⁹

Results in Table 2 show that cities with growing Twitter chatter about fracking have a greater likelihood of organizing events, an important finding that suggests local activists used Twitter as an organizing tool. ⁴⁰ Results in this table also show that population size and location in any region directly above the Marcellus Shale (with or without existing wells) increase the likelihood of organizing events in a city. We also find a very strong effect of the overall scale of the nonprofit sector, suggesting that a city’s overall civic capacity represents a substantial influence. Indeed, our review of the field of organizations active in making anti-fracking claims illustrates that the majority of these groups are ad hoc coalitions, built from preexisting blocs of citizens, which have sprung up in recent years to fight against this contested activity. The presence of environmental organizations and proximity to water contamination sites are not significant influences on anti-fracking mobilizations. ⁴¹ Proximity to wells and a greater share of Democratic voters decrease the likelihood of organizing events; both are surprising findings. However, it is also clear that individuals who live in close proximity to fracking often benefit materially from the activity or sign NDAs. The finding on partisanship likely reflects the property-rights orientation of many conservative landowners. Finally, local employment in the oil and gas industry increases the likelihood of mobilization—such workers are often a source of resentment in communities—as does location in Pennsylvania or Ohio.

Moving to our findings on local moratoria, our primary finding in Table 3 is that the anti-fracking mobilizations—which were, in part, motivated by screenings of Gasland—had a significant influence on the passage of moratoria against fracking. Model 2 shows that local Gasland screenings had a significant (p < .05) direct effect on the adoption of bans; yet, this effect becomes nonsignificant at p < .05 (although it is significant at p < .10) in Model 3, when the mobilization variable is added. Therefore, we conclude that Hypothesis 3 is not supported but Hypothesis 4 is supported.

As mentioned earlier, we also estimated additional models (reported in the online supplement) investigating the effects of Gasland screenings on local mobilizations using a matched set of cities with and without screenings, given concerns of endogenous selection; the results add further support to the argument that screenings had a significant effect on mobilizations within the following two months. ⁴²

Another important finding is that geographic proximity to previous adopters increases the likelihood of adopting a ban, which suggests activists are influenced by neighboring communities. These effects are stronger if we assume that closer cities exert a stronger influence than do distant cities. ⁴³ Unlike mobilizations, for moratoria we do not find significant effects of local Twitter chatter. However, as with our findings on mobilization patterns, general civic capacity was a factor in the passage of local moratoria; the overall number of nonprofit organizations in a city was a significant factor in the passage of moratoria, whereas we find no effect of the presence of specific environmental nonprofits. As in the previous case, we expect that the population of nonprofit organizations represented a significant resource on which anti-fracking activists could build and represent themselves against the threat of a potentially risky new practice. We also find that areas with higher rates of unemployment were less likely to pass fracking moratoria, confirming the expectation that economically vulnerable areas would not want to cut off a potential source of investment. In contrast, areas with greater population size and income are more likely to pass fracking moratoria.

Discussion and Conclusions

This study makes a number of important contributions to the literatures on social movement outcomes, media, the environment, and political change. Our findings make clear that cultural artifacts created by actors allied with social movements have both national and local effects; in our case, they helped displace expectations that gas drilling is safe with an alternative frame of fracking risk. At the national level, we show that the documentary Gasland created a discursive opportunity, particularly in social media. An important and growing body of work highlights the benefits that accrue to activists when they find “discursive opportunities” to make their claims politically resonant, but earlier work does not examine in sufficient depth how discursive opportunities emerge and change. We demonstrate that Gasland’s nationwide release on HBO and nomination for an Oscar award contributed to the emergence of a more favorable set of discursive opportunities for the anti-fracking movement. Specifically, Gasland contributed not only to an increase in online public attention to fracking, but also to increased chatter about fracking on social media and to increased coverage of the topic in mass media. Moreover, Gasland influenced the social media discussion about fracking by supporting diagnostic frames related to water pollution and, to a lesser degree, supporting prognostic frames regarding local bans and moratoria. Yet, Gasland’s influence on mass media’s discussion about fracking was weaker: mass media discussion did not become more focused on water problems or on bans and moratoria.

Our study shows that the release of the Gasland documentary influenced public discourse on fracking by highlighting environmental risks and offering possible solutions. This influence was larger in the Twittersphere than in the mass-mediasphere, in part because mass media coverage was more “balanced,” giving equal weight to critics and proponents of fracking. ⁴⁴ We argue that using only mass media sources to understand discursive opportunities may lead scholars to wrongly conclude in the direction of a “false negative,” in which opportunities are overlooked. Social media discourses may, in many cases, be much more aligned with movement claims and preferences than what is represented in the newspaper or on television. Some of this divergence may be due to mass media’s balance norms (Boykoff and Boykoff 2004; Gans 1979; Tuchman 1978), or to elite public relations campaigns and other efforts by counter-movements to challenge ideas favorable to a social movement (Vasi 2011; Vasi and King 2012; Walker 2009, 2014).

Readers might express concern that we have investigated the impact of a documentary film that was perhaps unusual in its ability to generate social change. Indeed, we find it remarkable that Figures 1 and 2 both show periods when discussion and Internet searching of “Gasland” surpassed overall discussion/searching for “fracking” (on Google, YouTube, and Twitter). Hundreds of political documentary films are released each year, and most will not generate very much discussion or activism. Given such considerations, we believe our study is generalizable for understanding the impact of documentary films that gain widespread popular attention, and future researchers could investigate similar effects for analogous films like Blackfish, An Inconvenient Truth, and The Cove.

At the local level, we show that screenings of Gasland in different locations had an effect on the mobilization of local campaigns against the controversial practice of hydraulic fracturing; in turn, these local mobilizations made local policymakers significantly more likely to take action to ban the practice of fracking in cities across the Marcellus Shale states. Social movements try to use cultural artifacts, such as documentaries, to build support for their cause and win policy changes. These cultural artifacts are not merely epiphenomenal traces of movements’ organization-building efforts but are themselves key mobilizing tools that help movements address the problem of quiescence (Crenson 1971; Gaventa 1982; Roscigno 2011) and counter opponents. Other studies of social movement outcomes show the importance of cultural artifacts, but this study is among the first to provide systematic evidence of their impact on both mobilization and political outcomes.

Taken together, these findings suggest that Gasland opened up a new discursive opportunity for the movement, increasing sympathetic social media chatter. That opportunity, in turn, supported anti-fracking mobilization in local communities. We find that both Twitter chatter and Gasland screenings had direct effects on organized events. But when it came to actually passing local bans, neither Twitter chatter nor the Gasland screenings were directly influential. What mattered more were anti-fracking mobilizations and diffusion processes. Discursive opportunities were critically influential in setting the stage for influencing social movement outcomes. In this respect, our finding is consistent with bodies of work on both political and discursive opportunities, which find that these effects are indirect.

This study also makes clear the importance of civic capacity for a community’s ability to respond to threatened environmental changes. Other studies have found similar evidence regarding the siting of contentious facilities (Aldrich 2008; McAdam and Boudet 2012). Here we find that communities’ capacities for mobilizing and changing policy were very much influenced by the density of preexisting nonprofits. In addition, the presence of local activists who were active on social media increased communities’ capacity for mobilization. These factors mattered more than the density of environmental nonprofits, suggesting that anti-fracking activity needed to build from the overall civic capacity of an area. Of course, consistent with previous research, much of this NIMBY activism reflects a more civically advantaged community’s ability to resist unwanted land uses.

We encourage future researchers to investigate the effects of other documentaries, films, and consciousness-raising tactics. Given that most of these events do not take place outdoors, on streets or public squares, but indoors, in cinemas, public libraries, or college campuses, these events are less visible and, therefore, harder to study. However, further research is necessary to understand why only some documentaries lead to large-scale mobilizations or how these less visible tactics contribute to social change. We also encourage future researchers to investigate the role of industry counter-mobilization through grassroots efforts and public relations campaigns on behalf of natural gas interests, as this may also affect community mobilization processes, media discourse, and ultimate political outcomes. There is some initial evidence of these influences. First, it is clear that natural gas interests have been heavily mobilized in favor of fracking, through campaigns organized by trade associations such as America’s Natural Gas Alliance and third-party groups like Energy in Depth. Furthermore, the natural gas industry helped promote the documentary Truthland, which raised questions about the accuracy of Gasland, and the counter-documentary FrackNation, which attacked Fox’s evidence and credibility. Given the issue’s shift from a place of almost complete novelty to such an established and contentious position over less than a decade, fracking serves as an excellent example of how movements grapple with contentious energy issues and the environment.