Tangled Roots: Personal Networks and the Participation of Individuals in an Anti-environmentalism Countermovement

The British Columbia Environmental Movement

This study centers on the conflict over Clayoquot Sound, a relatively pristine area on the west coast of Vancouver Island, on the west coast of Canada. A dispute over logging in this area became the biggest environmental protest (to date) in Canadian history. About 1,000 people were arrested in 1993 for blockading logging roads to stop clear-cutting in this area (Tindall, Robinson, and Stoddart 2015). The countermovement described in this paper arose in response to the actions of the environmental movement (and government responses to it) in Clayoquot Sound and other nearby areas.

From a research design perspective, this study adopts a cross-sectional survey design to examine several theoretically motivated hypotheses, as opposed to being a case study design per se. Nevertheless, we believe it will be instructive to provide further details about “the case” upon which data were collected.

The province of British Columbia is home to a substantial proportion of the world’s remaining old-growth temperate rainforests. These forests are highly valued by a variety of stakeholders, including Indigenous peoples, nature lovers, and forest-industry workers. Starting in the 1980s, disputes arose over the plans of forestry companies to log old-growth rainforests on western Vancouver Island. During the late 1980s and early 1990s, environmental groups lobbied to have a number of these old-growth forests protected as wilderness areas. The companies holding timber licenses for these publicly owned forests voluntarily set aside some “token” display areas, but vigorously maintained their right to log these forests, and continued the contentious practice of clear-cut logging in many of the disputed watersheds. In 1990, a partial resolution was brought about for one contentious area, the Carmanah Valley, when the British Columbia government declared part of the valley a provincial park. However, protests continued regarding other areas on the west coast of Vancouver Island (including Clayoquot Sound, as described above) and other parts of the province. Figure 1 provides a map of the geographical area that is the focus of this study.

OPEN IN VIEWER

Figure 1.

Map of Vancouver Island.

In response to pressure from environmentalists—or so it would appear—the provincial government took a number of actions (Tindall et al. 2015; Wilson 1998). Some of these actions include the establishment of a Commission on Resources and Environment (CORE), the creation of a number of new parks on Vancouver Island (including the entire Carmanah Valley), and the establishment of rigorous new legislation regarding forest practices.

The environmental movement has its base in the urban centers of Greater Vancouver and Greater Victoria. We focus on the segment of the environmental movement based in Victoria, as it is geographically closest to the disputes that are the main focus of this study (e.g., the Carmanah Valley, and Clayoquot Sound), many environmental social movement organizations have head offices or branch offices in Victoria, and it is the capital of British Columbia.

Greater Victoria, a largely middle-class service oriented economy, had more than 300,000 people at the time of the study. It is home to the government and to a university. Tourism is also an important part of the economy. There is practically no manufacturing. There are some fishers based there but beyond this there is little direct involvement in primary resource industries. Thus, the population is largely white collar and middle class. These characteristics are reflected in the membership of environmental organizations in the city (Tindall 2002).

Port Alberni and Share Our Resources

Port Alberni is a forestry-based community on central Vancouver Island (see Figure 1). It is also home to the countermovement organization that is the main focus of the present investigation: Share Our Resources of Port Alberni. In 1996, the population of Port Alberni was 18,468. Patricia M. Marchak, Scott L. Aycock, and Deborah M. Herbert (1999) summarize the community as one with a strong historical connection to forestry, a connection that made it prosperous. However, in the 1990s, with a downturn in the forest industry and accompanying loss of forestry sector jobs, Port Alberni was a community struggling to build a future.

The southern tip of Clayoquot Sound is about a 90-minute drive from Port Alberni. It is a visually spectacular area of fjords and mountains covered by ancient rainforest. A 1993 provincial government decision designated over two thirds of the area for logging. During the summer of 1993, the conflict over Clayoquot Sound became an issue of international importance.

These events strongly affected the residents of Port Alberni. In reaction to the creation of new parks and new forest policy that was implemented in the aftermath of the protests, and that was perceived to favor environmentalists over the forest industry and forest communities, a voluntary pro-forestry organization was formed: Share Our Resources of Port Alberni. This is one of many Share Groups that formed in the province to mobilize for the interests of resource-based communities (Doyle, Elliot, and Tindall 1997; Gutstein 2015; M’Gonigle 1989; Simon 1998). The Share movement is very similar to the Wise Use movement in the United States (Helvarg 1994; Rowell 2017), and several researchers have documented connections between the two movements (Gutstein 2015; Simon 1998).

Share Our Resources might be considered a “quasi-astroturf” group. It is an astroturf group (Lyon and Maxwell 2004) in the sense that it had assistance in its creation from corporate actors (Simon 1998), and from actors involved in the Wise Use Movement such as Ron Arnold (DeSmog, n.d.). However, we use the term “quasi” because the group really did exist, with numerous local members (some of whom we interviewed), and participated in protests rallies (which we attended). In other words, it was not an organization that was made up purely of public relation (PR) firm employees.

During the blockades in 1993, many Port Alberni forest workers and other residents had face-to-face interactions with the protesters at Clayoquot Sound. Furthermore, many Port Alberni forestry workers were laid off as a result of the reduced timber cut in Clayoquot Sound. Not surprisingly, relations between environmentalists and other community members in Port Alberni became tense.

M’Gonigle describes those who worked within the forestry profession as characterized by the traits of rationality, professionalism, and managerialism. He notes that these attributes contrast sharply with the foresters’ perceptions of the wilderness advocate:

Wilderness advocates were seen as: as irrational, driven by highly personal values or even spiritual passions, and appealing to the emotions of the citizenry. They are unprofessional, manipulating the citizen and the media, making exaggerated unscientific statements, and resorting to civil disobedience when their demands are not met. In contrast to the manager who must systematically balance the varied objectives which he or she is charged to implement, the preservationist is seen as a single-interest pressure group seeking to impose his or her particular values and objectives (ends) on the larger society. (M’Gonigle 1989:352)

Participants in Share or Resources were motivated to counter this perceived adversary. The Share Groups were the mirror images of the environmental movement groups of the time—for example, engaging in many of the same repertoire of activities, but in pursuit of opposite ends (e.g., making forests and other natural resources available for extractive economic exploitation). Activities included giving community talks, holding rallies, organizing protest demonstrations, lobbying politicians, writing letters to newspaper editor, participating in land planning round tables, and participating in media interviews, among other activities.

Networks and Movement Participation

Social network analysis is one approach to studying social structure. David B. Tindall (2002, 2004; Tindall and Robinson 2017) has developed an argument of how the structure of personal networks are related to level of ongoing social movement participation under conditions of low-medium cost/risk activism. The general thrust of the argument is that the greater one’s integration into a movement through network ties to other members of the movement (or network centrality): (1) the more often she or he will receive requests to participate in organization activities (ongoing network-based recruitment) and (2) the more opportunities she or he will have to discuss movement issues with others. Receiving requests to participate in organization activities and discussing movement issues and events with others are two examples of network-based processes that mediate the relationship between social structure and individual activism. This model asserts that network-based communication has both a direct and an indirect influence on activism. Information and knowledge partly underlie the positive relationship posited between frequency of communication and level of activism. The more often one discusses movement issues and events with others, the more informed and the more knowledgeable they will become about movement issues and events. As people become more knowledgeable about events and issues, they become more likely to participate in movement events.

Frequent communication about movement events and issues can have another effect upon potential and budding activists. Frequent conversations between individuals in small groups can serve as interaction rituals (Collins 2014) which help to shape and reinforce group identities. Hence, the more frequently one communicates about movement events and issues, the more highly she or he will come to identify herself or himself as a member of the movement.

Finally, as both social psychologists and social movement researchers have observed, group identification can serve as a basis for collective action. Identification is another “network-based process” variable that mediates between social structure and participation (Friedman and McAdam 1992).

The solid lines in Figure 2 illustrate the main direction of causation theorized by the argument here. However, the dashed lines also illustrate that (to some extent) the correlation between ego networks and activism is based on reciprocal causation. In some instances, people make ties to contacts that they meet at social movement events. As this study is cross sectional in design, we cannot separate social influence from social selection effects.

OPEN IN VIEWER

Figure 2.

A network centrality theoretical model explaining level of low risk/cost countermovement activism.

Characteristics of Personal Networks

There are many conceptions of social network structure. In this study, we utilize a few relatively simple measures of network centrality that can be applied at the level of the individual. Analyses that focus on individuals are often referred to as studies of personal networks or ego networks. Network centrality refers to the idea that individuals can be more or less in the middle of things in a network and the parallel idea that networks themselves can be more or less centralized. Early work on network centrality was focused on its implications for communication. The basic idea is that the more central an actor, the greater the degree of his or her involvement with others in a social network. The research described in this article focuses on personal or egocentric networks and thus utilizes measures of point centrality. The simplest and most intuitively obvious conception of centrality is one that is based on degree (Freeman 1978–1979). The degree of a point is calculated by counting of the number of other points that are adjacent to it, and with which it is in direct contact. Degree is an indicator of an individual’s potential communication activity, and thus theoretically influences social movement participation in a variety of ways.

The analyses presented here are organized around the following guiding hypothesis:

Network Degree and Network Range

In this analysis, we use the term degree to refer to the number of ties of the same type an individual has to others (e.g., number of ties to people in their own organization). We also use the concept range. Network range refers to the number of ties one has to different types of people (or social groups; Burt 1980). Degree is important for frequency of interactions, which in turn is important in the processes of information dissemination, social pressure, and attitude formation. Range is important for diversity of interactions; diversity of interactions is associated with the variation of information one receives, which in turn may be important for more complex cognitive processes such as identity formation. In measuring network range, we count the number of different movement organizations that respondents had personal ties to and memberships with.

For the degree measures, we investigate two different types: weak ties and strong ties to members of one’s organization.

Contesting Movement Ties

Doug McAdam and Ronnelle Paulsen (1993) identify a number of gaps in the literature on networks and social movement participation. One of these is a dearth of research on behavior outcomes resulting from the pressures of conflicting social ties:

Finally, and perhaps most important, the existing studies fail to acknowledge conceptually or treat empirically the fact that individuals are invariably embedded in many organizational or associational networks or individual relationships that may expose the individual to conflicting behavioral pressures. This weakness is due to all the well-known dangers of sampling on the dependent variable. Almost invariably, the studies of movement recruitment start by surveying activists after their entrance into the movement. But showing that these activists were linked to the movement by some preexisting networks does not prove the causal potency of that tie. No doubt there are many others who had ties to the movement but did not participate in its activities. We suspect one of the principal reasons for the failure of the tie to impel participation in these cases is the existence of other, perhaps more salient ties that are constraining involvement. But to date our lack of conceptual models of the recruitment process and the tendency to study activists after the fact of their participation has left the effects of these “multiple embeddings” unexamined. (Pp. 641–42)

The results to be reported later will allow us to examine some of these gaps.

Structural Hypotheses

Based on the insights of McAdam and Paulsen (1993), we have developed a set of structural (or social network) hypotheses. We hypothesize that network structure distinguishes countermovement participants from the general public. ² We also hypothesize that within-movement ties are positively correlated with activism and contending movement ties are negatively correlated with activism. Hence, we expect the following:

The above structural hypotheses constitute the core focus of the analyses that are to follow. (Each hypothesis is visually identified in Figure 2.) In particular, we will focus primarily on H3 and H4. (Other hypotheses are provided to give context and provide a systematic structure for the analyses to follow.)

H4 constitutes a novel line of investigation and, as such, merits further discussion. The main theoretical assertion is that if ego has a social tie to the contending movement (the environmental movement), this might be expected to constrain ego’s involvement in the anti-environmentalism countermovement due to counter pressures from alters and the social psychological mechanisms of balance theory (Heider 1958). One such mechanism involves adjusting one’s orientation to be more consistent with those of one’s alters. The present study uses cross-sectional data, so we can only examine the correlation between contending movement ties and countermovement activism and cannot easily assess causality. However, a few more words might be said here. First, we would expect this effect to occur if ties to the contending movement were preexisting and to be more powerful if the ties were strong. (But see Gould 2003 for an alternative point of view.) On the contrary, social ties can also be formed as a consequence of activism (e.g., see Tindall, Cormier, and Diani 2012), and such ties might also be correlated with level of activism. Thus, while we offer a specific hypothesis based on prior theoretical speculation, there are several possibilities in terms of empirical outcomes.

Cultural Considerations

In the analyses presented here, to guard against potential alternative explanations to the structural thesis we advance, we control for several key “cultural variables.” Below we provide theoretical rationale for the inclusion of these variables. A commonly offered explanation for people’s participation in social movements and other forms of protest is their adherence to particular values, attitudes, and/or ideologies (Klandermans 2015; Klandermans and Oegema 1987). Hence, we contrast our “structural argument” (McAdam 1986) with the potential explanatory importantance of value orientation, and identity.

Anthropocentrism/Biocentrism

A partially overlapping concept to the materialist/postmaterialist distinction is the anthropocentrism-biocentrism continuum, which describes an aspect of individuals’ relation to the environment (Larson 2010). People who argue that humans are part of nature, who question the right of humans to dominate and exploit nature, and who argue that nature has intrinsic value are often described as having a “biocentric” or “ecocentric” orientation (Catton and Dunlap 1978; Dunlap et al. 2000). People who argue that nature is a resource for humans to use as they please and that the primary consideration regarding our relation to nature should be based on utilitarian considerations are conversely described as being “anthropocentric.”

We expect that “biocentrism” is one of the values that underlies participation in the environmental movement, while “anthropocentrism” underlies participation in the anti-environmentalism countermovement. With regard to this distinction, a countermovement survey respondent offered the following written commentary on his questionnaire:

Forest policy seems to be driven by a biocentric view, leading to a diminished consideration of the economic well being of rural residents.

Because we are primarily interested in understanding participation in the countermovement, we will focus on the “athropocentrism” aspect of the continuum. As Figure 2 illustrates, we would expect anthropocentrism to be correlated with countermovement identification and countermovement participation.

More generally, Paul C. Stern et al. (1999) have developed a value-belief-norm (VBN) theory of movement support. They argue that individuals who accept a movement’s values and who believe that things they deem to be important are threatened, and have a sense of efficacy regarding their ability to protect or restore the things they value, will embrace an obligation (or personal norm) for pro-movement action that in turn makes individuals more likely to take actions to support a given social movement. While Stern et al. (1999) do not talk about countermovements in particular, we would expect a similar type of logic to describe the relationship between VBNs and countermovement participation (see also Dietz 2015).

While Stern et al. (1999) and other authors working in environmental sociology have provided theoretical rationales for a value basis for social movement participation, social movement and social network scholars have long argued (theoretically and empirically) that while movement congruent values and attitudes might be a necessary condition for participation, they are not sufficient. Even for individuals with movement-congruent views, what is critical are personal network ties to others involved in the movement (Klandermans and Oegema 1987; McAdam 1986). The present analysis examines both values and orientations (as control variables) and the effects of network ties on countermovement activism.

Defining Identity against the Environmental Movement

By definition, countermovement mobilization is mobilization against social change and/or social movement mobilization (Meyer and Staggenborg 1996). Qualitative findings from our survey, field research, and interviews indicated that countermovement participants were motivated to get involved as a reaction to the comments and actions of environmentalists (see also Lewin 2019). Based on these observations, we conjecture that identification with the environmental movement (strength of agreement that the label applies to them) is negatively associated with level of countermovement identification and participation (see Figure 2).

Data Collection

The current project incorporates a mixed-methods approach to data collection (interviews, survey data collection) and mixed methods of data analysis (qualitative and quantitative; Cresswell 2003). The first phase of the project involved a structured interview using open-ended questions with organizational representatives (of organizations involved in contention over forestry issues on Vancouver Island). The second phase involved a survey of the members of Share Our Resources of Port Alberni.

John W. Cresswell (2003) discusses the decisions a researcher needs to consider when undertaking a mixed-methods approach. He argues that researchers need to consider (1) the sequence of implementation, (2) the priority given to the qualitative versus quantitative component, (3) how the two different components are integrated, and (4) the theoretical perspective (whether explicit or implicit). In terms of the larger vision of this project, the sequence and the emphasis was qual QUAN—although the current paper will focus primarily on the quantitative analysis of the last phase.

Information obtained through the interviews with organization representatives helped with the content design of some of the items for both the self-administered questionnaire (such as questions about particular opinion items) and interpretation of the quantitative results (such as interpretation of the effect for contending movement ties).

Because of space limitations, we only provide an in-depth description of the quantitative part of this project. In terms of theory, we emphasize a version of social network theory that stresses social influence (e.g., the effect of network ties on movement participation), but we also talk about the possibility of social selection effects (the formation of social ties based on the attributes of individuals).

This article is based on three sources of data collected in the 1990s: (1) a self-administered questionnaire distributed to a probability sample of members of “Share Our Resources” (an anti-environmentalism countermovement organization) in Port Alberni (N = 129); (2) a telephone interview survey with a probability sample of the Port Alberni general public (N = 100); and (3) a self-administered questionnaire mailed to probability sample of members of three formal environmental movement organizations in the Greater Victoria Area. In addition, interviews were conducted with members of the environmental movement, countermovement, and community representatives in Port Alberni, and further data were collected through observational techniques and collection of available documents.

The main focus of this study is upon the “Share Our Resources” data (source 1) and thus more details will be given about these. Measures for the data collected for the Port Alberni general public (source 2) were either identical to those used in the Share Our Resources questionnaire or based on a subset of items (in the case of some indexes—such as level of activism). (The “general public” sample is a probability sample of the general public. As noted earlier, it theoretically includes respondents with varying views and behaviors, including, possibly, individuals who are pro-social movement, pro-countermovement, or neutral. For more information on the Port Alberni general public survey see Tindall and Robinson 2017.) Measures for the environmental movement are identical or nearly identical measures to those administered to the Share Our Resources sample. Where differences exist, it is primarily in the wording of items—to make them relevant to environmental movement members. (For more information on the environmental movement survey see Tindall 2002).

Missing Values

For the countermovement sample, missing data in the form of nonresponses existed for small proportions of cases for the variables included in the forthcoming analyses. The SPSS multiple imputation procedure was undertaken to impute missing values (IBM Software Group 2016). In some of the tables reported in the results, findings are provided based both on the original data (where cases are dropped if there are missing values) and on the multiply imputed data which contains all the cases. ⁴

Level of Activism

The main dependent variable of interest is level of activism. In the questionnaire, respondents were asked to indicate which types of activities, from a set of movement-related items, they had ever participated in ⁵ (items are listed in Appendix Table 1). An index of activism was constructed from these items. More details are given in Appendix Table 1. Descriptive statistics for key variables are given in Tables 3 and 4.

Network Variables

Contending Movement Ties (Range of Ties to Environmental Organizations)

Members of the countermovement organization sample and the general public sample were provided with a list of environmental organizations and were asked for each organization if they were a member of the organization, and whether they knew anyone who belonged to the organization. Two variables were created: contending movement ties and contending movement organization memberships.

Level of Identification

Level of identification for the countermovement is measured by asking respondents the following: how strongly they identify themselves as a member of Share Our Resources and how strongly they think that others identify them (the respondent) as a member Share Our Resources. Responses to these items were combined into an index.

Materialist Values Index

Motivated by Inglehart’s (1990) work, respondents were asked to rate the relative importance of a set of materialist and PMVs. The items used here were borrowed from the Canadian Election Study. Past research by the first author revealed that respondents found ranking the items to be quite taxing. Thus, the item was revised and responses were based on rating on a 10-point scale (with 1 indicating greater importance). For countermovement respondents, an index of support for materialist values was constructed by adding together responses for a subset of the materialist value items.

Anthropocentrism Index

Respondents were asked to rank the relative importance of six forest values. This questionnaire item was designed for and previously used in a Canadian Forest Service survey. Two of the items were “anthropocentric” in nature (“a place for recreation and relaxation,” “a source of economic wealth and jobs”). For countermovement respondents, an index of “anthropocentrism” was constructed by summing responses to these two items. (For environmental movement respondents, an index of biocentrism was constructed from the other four items.)

Analyses

A series of descriptive bivariate and multivariate analyses follow. The data were examined for statistical assumptions and regression diagnostics were run. Based on these, it was decided to run ordinary least squares (OLS) regression/multiple regressions for the analyses involving activism as a dependent variable, and Poisson regression for the regression where contending movement ties was the dependent variable. For the former multiple regressions, several cases were removed from the analyses because of the existence of substantial outliers. Analyses were run with and without the outlying cases, with no change in the substantive findings. The results reported here use data with the outliers removed, to remove the bias of these outliers.

The main analyses presented here are OLS multiple regression and Poisson regression. These methods are appropriate for the data examined here. It should be noted that exponential random graph models (ERGMs) are often employed for analyzing social network data because tie dependencies across cases in whole network data violates the assumption of the independence of cases that underlies normal regression (Lusher, Koskinen, and Robins 2013). Because the data analyzed here are from random samples of ego networks, this concern does not apply, and procedures such as OLS regression are appropriate.

Explaining Level of Identification and Level of Activism Using Cultural Variables (among Countermovement Members)

Table 1 shows regression results for models using the “countermovement cultural variables” to predict level of countermovement identification and level of countermovement participation. Results for level of countermovement identification show some limited support for the cultural basis of countermovement mobilization discussed earlier and illustrated in Figure 2. Respondents who more highly rated materialist values identified more highly as a member of the countermovement organization. Similarly, respondents who more highly ranked “anthropocentric forest values,” more highly identified as a member of the countermovement organization. Finally, the negative coefficient for level of environmental movement identification means that the more strongly a respondent disagreed that they were an environmentalist, the more strongly they identified as a member of the countermovement organization. While the variables are not significant at <.05 in the multiple regression, materialist values and countermovement identity are significant in zero-order correlations, and the p value for anthropocentrism = .054.

OPEN IN VIEWER

Table 1.

Multiple Regression Analysis Predicting Level of Identification and Level of Activism Using Cultural Variables as Independent Variables for Countermovement Members.

Independent variables	DV: Level of countermovement identification	DV: Level of countermovement activism
Values
Materialist values index	−.19	.07
Anthropocentrism index	−.16	−.21*
Identification
Level of environmental movement identification	−.13	.06
R 2	.09*	.05
Adjusted R2	.06*	.02
N	106	106

Note. Standardized regression coefficients are shown in the table. DV = dependent variable.

Two-tailed tests of signficance provided.

*

p ≤ .05. **p ≤ .01. ***p ≤ .005.

Results reported in column 2 of Table 1 show that only ranking of anthropocentric forest values is significantly associated with level of countermovement activism. However, as results to be shown later will demonstrate, level of identification with the countermovement is one of the strongest predictors of level of countermovement activism—and thus materialist values and environmental movement identification have small indirect effects on level of countermovement activism through their correlations with level of countermovement identification. In sum, in terms of the measures explored here, there are only small effects for the countermovement culture variables on identification and activism. For both equations, the amount of variation explained is small. Thus, explanations for countermovement identification and activism need to be sought elsewhere. The analysis provided here will focus primarily on activism.

Explaining Activism: Comparing Models for the Countermovement and the Environmental Movement

As noted earlier, Tindall (2002, 2004) has presented a network model of individual participation in low cost/risk social movement activism among members of the wilderness preservation stream of the British Columbia environmental movement. In general, he argues that individuals who are more integrated into the movement through their ties with other movement members will receive more requests to participate in ongoing activities, communicate more frequently about movement issues, identify more highly with the movement, and be more active.

Table 2 compares results for members of the anti-environmentalism countermovement organization with members of the British Columbia environmental movement (methodological details for the latter are described in Tindall (2002)). (Columns 3–5 provide findings from the multiply imputed data sets.) We will focus on the network and network-process variables. The results for members of the environmental movement (column 1) support the general argument about networks and participation provided above. Results for countermovement members also generally support this argument and mirror findings for the environmental movement. In most instances, the pattern is the same for the countermovement as for the environmental movement, but in some cases, effects are nonsignificant for the countermovement because of the smaller sample N.

OPEN IN VIEWER

Table 2.

Multiple Regression Analyses with Activism as the DV for Countermovement and Environmental Movement.

Independent variables	Environmental movement DV: Activism	Countermovement DV: Activism original data	Minimum	Maximum	Pooled significance
Control variables
Length of membership	.05	.20*	.14	.27	*
Age	−.18***	−.07	−.03	.00
Gender	.02	.13	.10	.14
Education	.06	.11	.09	.17
Values
Materialist/postmaterialist values index	.00	−.07	−.05	.06
Anthropocentrism/biocentrism index	.07	.08	.04	.13
Network structure variables
CMO/EMO weak ties	.20***	.15	.10	.12
CMO/EMO strong ties	−.03	.05	.05	.10
Range of CMO/EMO ties	.31***	.19*	.16	.21	*
Range of CMO/EMO memberships	.23***	.03	.08	.11
Network-based processes
Ongoing network-based recruitment	.11**	.05	.06	.10
Frequency of communication	.13***	.16	.12	.16
Level of movement identification	.24***	.24*	.13	.20	†
R 2	.66***	.38***	.36	.41	***
Adjusted R2	.65***	.27***	.27	.34	***
N	241	85	126	126

Note. Standardized regression coefficients are shown in the table. DV = dependent variable; CMO = countermovement organization; EMO = environmental movement organization.

Two-tailed tests of signficance provided.

†

p ≤ .10. *p ≤ .05. **p ≤ .01. ***p ≤ .005.

In comparing the two network degree measures (weak ties and strong ties), we find that in both cases, the weak ties measure is a stronger predictor of activism. ⁶ Of the two network range measures, number of within-movement organization ties is significant for both movements, but number of organization memberships is only significant for the environmental movement. However, the nonsignificance of this effect for the countermovement may in part be an artifact of the measure. There were fewer organizations that were part of the countermovement (four) relative to the environmental movement (seven), and the amount of possible variation for this variable is therefore limited.

With regard to the network-based process variables, both are significant for the environmental movement. Both are positive though nonsignificant for the countermovement (however, the coefficient for ongoing network-based recruitment is very small).

Finally, level of movement identification is positive for both groups. It is the strongest predictor of activism among countermovement participants, and the second strongest predictor among environmental movement participants.

Overall, there is a similar pattern of results for the two groups, generally supporting a network model of individual participation (Tindall 2002). As noted, several coefficients for the countermovement sample are similar in sign to the result for the environmental movement sample but are nonsignificant. (It should be noted that in reduced models, these coefficients are similar in magnitude and signs, and are significant.) Finally, when network factors are controlled, the “culture variables” (materialist/PMVs, anthropocentrism) are nonsignificant. This supports the arguments of scholars that network factors are better predictors of social movement participation than values and attitudes (e.g., Klandermans and Oegema 1987).

While these results suggest that similar processes are at work in both the countermovement and the environmental movement regarding within-movement ties, several important network-related questions remain, as suggested by McAdam and Paulsen (1993): (1) How do the personal networks of countermovement participants compare with the general public? (2) Do similar within-movement tie effects on participation effects exist for the general public? (3) How do contending movement ties (e.g., ties to environmental movement members) affect participation among countermovement participants? To answer these questions, it is necessary to explore the characteristics of personal networks among countermovement participants more closely and to examine data for the general public.

Comparing the Countermovement and the Port Alberni General Public: Activism and Network Ties

Tables 3 and 4 provide descriptive statistics for the countermovement and for the general public within Port Alberni, the town that the countermovement organization was based within. Table 3 provides means and standard deviations for activism and for selected network variables. Unsurprisingly, countermovement participants were somewhat more active than the general public. There is relatively little difference in terms of range of ties to pro-forestry countermovement organizations—though on average, countermovement participants have slightly more ties to countermovement organizations (H1).

OPEN IN VIEWER

Table 3.

Comparing the Countermovement and Port Alberni General Public: Activism and Network Ties.

Activism and Social Network Ties Measures.	Countermovement	General public
Activism (short list)	4.74	2.38
	(1.97)	(2.15)
Range of ties to countermovement organizations	1.87	1.49
	(0.90)	(0.92)
Range of memberships to countermovement organizations	1.34	0.49
	(0.49)	(0.72)
Contending movement ties (range of ties to people in environmental movement organizations, short list)	0.76	0.63
	(1.42)	(1.24)
Contending movement memberships (range of memberships in environmental organizations, short list)	0.01	0.09
	(0.09)	(0.35)
N	129	100

Note. The values represent means and standard deviation (in parentheses).

OPEN IN VIEWER

Table 4.

Comparing the Countermovement and the General Public: Weak Ties and Strong Ties.

Weak Ties and Strong Ties Measures	Countermovement (%)	General public (%)
Weak tie to countermovement organization (Share Our Resources)	68.2	35.0
Strong tie to countermovement organization (Share Our Resources)	44.2	12.0
N	129	100

Note. Percentage with at least one tie.

Countermovement organization members belong to a greater number of countermovement organizations. This is partly due to the fact that everyone in the countermovement sample belongs to the countermovement organization under investigation. While this seems unsurprising, it has important potential consequences. Theoretically, having even one countermovement organization membership (vs. none) should affect identification and participation in the countermovement.

In the last two rows of Table 3, two new “contending movement network” variables are introduced, namely, range of ties to people in environmental organizations and range of environmental movement organization memberships. There is very little difference here. Countermovement respondents are slightly more likely to have a tie to an environmental organization than the general public. Members of the general public are very slightly more likely to have an environmental organization membership—though as the table indicates, very few people in either sample belonged to an environmental organization. Our within-movement structural hypothesis is supported: Countermovement members are somewhat more integrated into the movement through their range of ties (H1) to, and memberships with, countermovement organizations. However, our structural hypothesis regarding contending group ties (H2) is not supported—it is not the case that the general public has substantially more connections to the environmental movement.

In sum, in Table 3, countermovement participants are more active, but aside from countermovement organization membership, the network characteristics of the two samples are fairly similar. So the question remains, what explains the difference in participation?

Table 4 provides the percentage of respondents in each group who had a strong tie and who had a weak tie to the countermovement organization. Percentages are given instead of means because only dichotomous data were collected for the general public. (Data for the countermovement respondents were collected at the ratio level but converted here into dichotomous form.) Results provided in Table 4 provide a possible answer to the question posed above. Countermovement members are more likely to have a weak tie to the countermovement organization (68.2 vs. 35 percent) and more likely to have a strong tie to the countermovement organization (44.2 vs. 12 percent). This finding provides further support for the structural hypothesis that countermovement members are more integrated (than are the general public) through their within-movement ties (H1).

In sum, in Tables 3 and 4, we have identified some similarities and differences between countermovement participants and the general public regarding activism and personal network characteristics. Two questions remain: (1) Are there network effects on participation for the general public? (2) What effects do contending movement ties have?

Explaining Level of Activism Using Different Network Variables Predictors (Countermovement and Port Alberni General Public)

Table 5 provides a series of regression models using different network variables to predict level of activism. Column 1 contains a model for the general public, columns 2 and 3 contain models for the countermovement sample. In columns 1 and 2, modified versions of the dependent variable and contending movement ties are used to make the results directly comparable. (The modified variables consist of a subset of items—as the general public was provided with short lists of activities and environmental organizations in their interview.) Column 3 of Table 5 (for the countermovement sample) includes the “full” activism and the “full” variable indices. Finally, weak and strong ties are included only in the model presented in column 3, as ratio data for these variables were available only for the countermovement sample. ⁷

OPEN IN VIEWER

Table 5.

Multiple Regression Analysis Predicting Level of Activism Comparing Ties to Environmentalists with Ties to the Countermovement (for Share) (Using Standardized Regression Coefficients).

Independent variables	General public (port Alberni) (short list index) a	Countermovement (short list index) a	Countermovement (activism)
Network structure variables
CMO/EMO weak ties	—	—	.19*
CMO/EMO strong ties	—	—	.07
Range of CMO/EMO ties	.39***	.10	.19*
Range of CMO/EMO memberships	.35***	.21**	.19*
Contending movement ties (range of ties to people in environmental organizations) b	−.00	.25***	.34***
R 2	.36***	.17***	.32***
Adjusted R2	.34***	.14***	.29***
N	97	126	126

Note. CMO = countermovement organization; EMO = environmental movement organization.

a

The “Short List” index contains a subset of items included in the full activism index. Use of this measure allows direct comparison between Countermovement and the Port Alberni general public.

b

The models in columns 1 and 2 use a “range index” based on a shortened list of environmental organizations.

Two-tailed tests of signficance provided.

†

p ≤ .10. *p ≤ .05. **p ≤ .01. ***p ≤ .005.

Results in column 2 for the countermovement show that all three range measures are positively associated with level of activism (though range of countermovement ties is nonsignificant). That the countermovement ties/memberships measures should predict activism agrees with theory (H3). However, the fact that the contending movement ties effect is positive and significant is surprising (H4). The literature suggests that such ties should have a moderating effect on activism and thus the coefficient should be negative (e.g., having both within-movement and contending movement ties provide competing pressures).

Results in column 1 for the general public show strong positive effects for both of the countermovement network range variables (H3). This finding is theoretically consistent but is novel in the sense that previous work on networks and social movement participation has focused largely on activists to the neglect of nonactivists (McAdam and Paulsen 1993).

Results for the general public (column 1) show no effect for contending movement ties (H4). This is a dramatic contrast with the results for the countermovement sample. Interpretation of this result will be undertaken in the discussion section. Interestingly, a greater percentage of the variance is explained for the general public than for the countermovement. This may be due to the fact that there was greater variation among the public sample for activism (see Table 3).

In column 3, the full version of activism and environmental organization ties (contending movement ties) measures are used and both weak and strong ties (to the countermovement organization) are also included in the equation. Here, the weak ties effect is positive and significant. Both countermovement network range measures are positive and significant (H3). Finally, number of contending movement ties remains positive and is again the strongest predictor of level of activism (H4). Once again, the findings regarding the within-movement (countermovement) tie variables (H3) replicates past research (Tindall 2002; Tindall and Robinson 2017), but the finding regarding contending movement ties (environmental movement ties) is novel and requires additional investigation (H4).

Explaining Level of Activism Using “Countermovement” Variables

In an effort to further examine the finding regarding environmental movement ties for the countermovement sample, Table 6 revisits the potential effects for the countermovement cultural variables illustrated in Figure 2, and also includes contending movement ties, and total countermovement organization ties (e.g., both weak and strong ties). (Results from the imputed data sets are provided in columns 3–5.) In column 1, we see that contending movement ties (ties to people in environmental organizations) is, by far, the strongest predictor of level of activism.

OPEN IN VIEWER

Table 6.

Multiple Regression Analysis Predicting Level of Activism Using “Countermovement” Variables (for Share) (Using Standardized Regression Coefficients).

Independent variables	Countermovement Model 1	Countermovement Model 2	Minimum	Maximum	Pooled significance
			Original data	Original data
Values
Materialist values index	.15 †	.15 †	.07	.12
Anthropocentrism index	.17 †	.13	.12	.15	†
Network structure variables
Contending movement ties (range of ties to people in environmental organizations)	.43***	.41***	.41	.41	***
Total countermovement ties (control)		.24**	.24	.25	***
Level of movement identification
Level of environmental movement identification	.04	.03	.01	.06
R 2	.23***	.28***	.28	.30	***
Adjusted R2	.20***	.25***	.25	.27	***
N	110	110	126	126

Two-tailed tests of signficance provided.

†

p ≤ .10. *p ≤ .05. **p ≤ .01. ***p ≤ .005.

The question arises as to whether the effect for environmental movement ties is merely a function of network size. For example, those with larger personal networks participate more, but they also have a greater number of contending movement ties—perhaps the contending movement ties effect for the countermovement sample is spurious (e.g., a gregariousness effect). Thus, in column 2, total countermovement organization ties is added to the model. While total number of ties is positive, significant, and explains an additional 5 percent of the variation in level of activism, we see that the coefficient for environmental organization ties remain positive, significant, and changes only very slightly in its magnitude. It remains the strongest predictor of level of activism.

Explaining Level of Activism: Full Model with “Countermovement” Variables (Countermovement Members)

Table 7 provides a final regression model with activism as a dependent variable to examine the contending movement ties effect. Here a full model is presented including the countermovement culture variables and all of the relevant network structure and network-process variables. (Columns 2–4 provide results from the imputed data sets.) In this final model, we see that while its coefficient is somewhat reduced compared with Table 6, range of contending movement ties remains the strongest predictor of level of activism (H4). The overall model explains a substantial amount of variation by social science standards. (The adjusted R² varies between .33 and .36 for the imputed models.)

OPEN IN VIEWER

Table 7.

Multiple Regression Analysis Predicting Level of Activism “Full Model” Using “Countermovement” Variables (for Share) (Using Standardized Regression Coefficients).

Independent variables	Original data	Minimum	Maximum	Pooled significance
Control variable
Length of membership	.19 †	.14	.25	†
Age	−.12	−.08	−.05
Gender	.10	.09	.12
Education	−.00	.02	.09
Values
Materialist/postmaterialist values index	−.01	−.00	.08
Anthropocentrism index	.12	.07	.15
Network structure variables
CMO/EMO weak ties	.17	.11	.13
CMO/EMO strong ties	−.02	.01	.04
Range of CMO ties	.14	.08	.12
Range of CMO memberships	−.01	.07	.10
Contending movement ties (range of ties to environmental organizations)	.32***	.26	.31	***
Network-based processes
Ongoing network-based recruitment	.11	.11	.14
Frequency of communication	.10	.06	.09
Level of movement identification
Level of countermovement identification	.20 †	.09	.15
Level of environmental movement identification	.00	−.01	.03
R 2	.46***	.42	.46	***
Adjusted R2	.33***	.33	.39	***
N	84	126	126

Note. CMO = countermovement organization; EMO = environmental movement organization.

Two-tailed tests of signficance provided.

†

p ≤ .10. *p ≤ .05. **p ≤ .01. ***p ≤ .005.

In Figure 2, while we identified the effect of contending movement ties as one of the main hypothesized effects on level of countermovement activism, we also noted that theoretically there should be a reciprocal effect of activism on contending movement ties. Thus, in Table 8, we provide a Poisson regression examining the statistical predictors of contending movement ties. Our main focus here is on the effect of level of countermovement activism on contending movement ties, but we also control for several other key explanatory variables.

OPEN IN VIEWER

Table 8.

Poisson Regression Analysis Predicting Contending Movement Ties (Poisson Regression Coefficients).

Independent variables	Original data (SE)	Minimum	Maximum	Pooled significance
Intercept	−7.25***	−6.35	−7.33	***
	(1.23)	(1.00)	(1.04)
Length of membership	0.01	−0.00	0.05
	(0.04)	(0.04)	(0.03)
Gender	−0.47	−0.56	−0.27
	(0.33)	(0.27)	(0.26)
Age	0.03 †	0.01	0.02	†
	(0.03)	(0.01)	(0.01)
Education	0.29***	0.26	0.32	***
	(0.06)	(0.05)	(0.05)
Materialist values index	−0.28***	−0.22	−0.13	†
	(0.11)	(0.08)	(0.08)
Anthropocentrism index	−0.06	−0.05	0.02
	(0.11)	(0.09)	(0.09)
Identification with countermovement	0.10	0.15	0.19	***
	(0.07)	(0.05)	(0.06)
Level of countermovement activism	0.11***	0.11	0.14	***
	(0.04)	(0.03)	(0.03)
χ2	69.59***	98.31	114.09	***
N	90	129	129

Two-tailed tests of signficance provided.

†

p ≤ .10. *p ≤ .05. **p ≤ .01. ***p ≤ .005.

In Table 8, as Figure 2 suggests, there is a significant positive effect of countermovement activism on contending movement ties. There are also significant effects for education and identification with the countermovement and marginal effects for age and the materialist values index. We will talk about these findings in the discussion.

Within-movement Ties and Activism

When examining within-movement ties, we observe similar network tie (H3) and network process effects for both members of environmental movement organizations, and members of an anti-environmentalism countermovement organization (Table 2; though the weak ties effect is only significant for the larger, environmental movement sample).

This study is relatively novel in that it provides a comparison between members of a countermovement and members of the general public. As expected (H1), countermovement members are more integrated into the countermovement than are members of the general public. However, contrary to expectations (H2), the general public of Port Alberni are not more connected to the environmental movement (Table 3).

Unsurprisingly, countermovement members are more active than are members of the general public (Table 3). Results suggest that one reason for this may be the fact that countermovement members are more likely to have both weak ties and strong ties (H1) to other members of the countermovement organization (see Table 4), and processes of influence associated with this greater level of integration explains activism.

Contending Movement Ties and Activism

The most interesting and surprising finding of this study concerns contending movement ties (H4). For countermovement members, the strongest statistical predictor of their level of participation in countermovement activities was their range of contending movement ties (connections to people in environmental organizations; see Tables 5 to 7). This is theoretically interesting because the existing literature suggests that having ties to people who oppose the movement will have a negative effect on activism (McAdam and Paulsen 1993), but just the opposite was observed. Also of interest is the contrasting finding for the general public—there is no contending movement ties effect (H4, see Table 5). The rest of the discussion will focus primarily on trying to understand these novel findings.

The contending movement ties effect is based on the number of environmental organizations where ego knows someone (either a weak ties or a strong tie). Results show, based on further post hoc analysis, that these ties are overwhelmingly weak. For the countermovement sample as a whole, there are an estimated 103 weak environmental organizational ties and only seven strong environmental organizational ties (these data were not presented earlier). Thus, this is primarily a measure of weak ties (Granovetter 1973) to environmentalists.

It might be reasonably argued that the “counter pressure” effect of a contending movement tie (stated in H4) should only occur for strong ties. Our response is that we would expect a much stronger “counter pressure” effect for strong contending movement ties. However, it seems plausible to expect that there may still be a small negative effect for weak contending movement ties (Granovetter 1973). Such ties are more likely to serve as bridges, and as such, might be sources of novel information. In some instances, information might result in persuasion. Furthermore, the criticism about the fact that we might expect a bigger effect for strong ties than weak ties does not explain why we observed a moderately strong positive effect for contending movement ties—as we would expect the effects to be negative.

Our field research indicated that relatively few people in Port Alberni belonged to environmental organizations. In fact, in terms of local environmental organizations, the main local group was comprised of about 20 members. ⁸ This contrasts with a membership of well over a thousand for the local countermovement organization. In terms of the general public Port Alberni survey data, 7 percent of the population belonged to an environmental organization (of some type—most of these are regional organizations rather than local ones), while 40 percent belonged to a countermovement organization. ⁹

What this implies is that while there are some local environmentalists active in the community, many (perhaps most) interactions that countermovement members had with environmental movement members is with nonlocal environmental movement members. If most environmentalists are not from the local community, where do interactions with environmentalists take place? We know from our interviews and from the survey data that many members of the countermovement participated in land planning processes in one way or another. Some sat on advisory boards. Others participated in community-based resource-planning meetings. These are some relatively important contexts in which countermovement participants met and interacted with environmentalists.

While there is a lot of anger and resentment about environmentalists (e.g., McCarthy 2002), the nature of movement-countermovement interactions are complex. Even central countermovement combatants expressed positive effects of such interactions on occasion. For example, one countermovement respondent stated on their questionnaire:

I got involved because of the outlandish comments and actions of the preservation groups. I also got involved because of the seemingly indifference and lack of action from M&B [the major forestry company in the community]. I became alarmed because if the preservationists weren’t checked they could wreck havoc on forest dependent communities. . . . Not all the preservationists’ efforts were negative. They have taught many, including me, to tread more softly on the earth.

Another context in which environmentalists and countermovement members interacted was at protest demonstrations and blockades. For example, for most of 1993, there was an environmental movement led blockade of an important logging road about 45 minutes outside of Port Alberni. Over the course of the year, many countermovement demonstrations were held at the blockade. Television coverage showed heated conflicts on the blockade line. A further example is provided by a 1994 pro-forestry rally that was held on the lawns of the provincial legislature attended by about 20,000 forestry workers and countermovement supporters. Some environmental activists also attended this demonstration and were the object of physical and verbal aggression.

In our interviews, local environmental movement informants told us quite a lot about their personal relationships in the community. Activists witnessed many of their ties with others in the community dissolve as a result of the conflict over old-growth forests in the area. Some countermovement respondents expressed that they had had similar experiences. Relatedly, many countermovement respondents complained that their community was being adversely affected by environmental movement outsiders. For example, one respondent provided the following written observation on her questionnaire:

The year of the Clayoquot Protest was one of the most unpleasant in my life. The self-righteous rhetoric and preaching from urban and European environmental missionaries was oppressive. The emotional polarization on the “save our trees” issue even impacted on long term friendships. For a while we successfully relied on humour to deflect some of the oh-so-serious missives and barbs aimed societally and personally. But the sense of attack didn’t abate. [I] felt drawn into the fray, which resulted in feeling angry and personally attacked. [I] didn’t like the feeling—it was unhealthy and counterproductive, so [I] tried to protect myself by establishing some distance from the issues and rhetoric . . .

In talking about relationships in the community, one countermovement interview respondent said,

I think it’s been very stressful for the community and for us that have been involved. And I was involved in the Clayoquot protest. It took about three years to get nowhere, and so in the CORE process I think it’s been stressful. Before the Province put the plan in place a few years ago with [Premier] Mike Harcourt, we had a Clayoquot process in place to try and work out what’s going to happen with Clayoquot, and we had people from all various areas working. I think we worked with it for about 3 years and really didn’t get anywhere. People who were entrenched in their position. All these processes took up an enormous amount of time.

A number of interviewees expressed their frustration about people from outside the local area and their perceived intransigence on environmental compromise. For example, one countermovement interview respondent commented:

Well, right now, I’m at a point in my life where I just say that, like you know, like I’ve been fighting this thing since 1987, I’ve been trying to get some resolve and basically the resolve that we’ve been trying to get from sharing resources, you know that’s why they’re called, “shared” resources, we just want to share out there, and unfortunately [there] is no sharing, you know there never can be sharing until people in British Columbia, especially in the urban communities realize that we have to have both. Environmentally, we have to have jobs. I just don’t’ think that we can carry on the way we’re going, because everybody loses out.

Countermovement interview respondents emphasized the roles and actions of nonlocal environmental movement protesters. For example, in talking about civil disobedience, the above interviewee asserted,

To continue the protest is breaking the law. And our courts have been way too lenient on people breaking the law- way too lenient- if you want to disobey a court injunction, then you should pay the price. If you are from out-of-state, or you’re out of Province, and you are not a Canadian citizen, and you protest and you break the law, then all the rights and privileges of coming into this country should be cancelled. And that’s my basic answer to that.

In their written responses to the questionnaire, others added,

Environmental protests are usually directed and led by people who reside outside the area of concern. Consequently they are not directly affected by the result of their action.

Those who live in Victoria or Vancouver should maintain their own communities. Let them look after their own backyards.

It appears that part of the “environmental ties effect” is resentment about outsiders who are adversely affecting the lives of local community members. This observation dovetails with the findings of Thomas Dunk (1994) on forestry workers in northern Ontario, and James McCarthy’s (2002) analysis of the Wise Use movement, and Lewin’s (2019) analysis of a coal mining community.

Contending Movement Ties as an Outcome of Activism

Post hoc analysis of the contending movement tie effect (ties to environmental organizations) suggests that this result is likely due—in part—to reciprocal influence (see Table 8 and Figure 2). In social network analysis terms, there are potentially both social selection and social influence effects (Snijders, Van de Bunt, and Steglich 2010). While within-movement ties draw some people into activism, and some social ties to the contending movement precede countermovement activism, those who are more active in the countermovement develop ties to environmentalists in the context of their activism (e.g., participating in multistakeholder land use planning processes, interacting with adversaries at protest demonstrations and blockades). Our analysis reported in Table 8 supports this view. On the contrary, it seems plausible that once countermovement activists make these ties, the network processes described earlier are reinforced. Interacting with environmentalists in conflictive situations reinforces the commitment of countermovement participants to the cause (level of countermovement identification being one proxy for this). More generally, these observations are consistent with Lewis Coser’s (1956) analysis of group conflict wherein he argues that, through group conflict, “groups become increasingly aware of their specificity,” establish boundaries between themselves and the outside/other groups, and become conscious representatives of the group (p. 115). This strengthens internal group cohesion and in social movements may facilitate mobilization.

With regard to considering contending movement ties (ties to people in environmental organizations) as an outcome variable, we also find effects for materialist values, age, education, and identification with the movement. The effects for materialist values and age are quite weak and may be the result of random variation. The effects for education and identification are stronger and are theoretically interesting. Regarding the latter, it makes theoretical sense that those who are most strongly committed to the movement in terms of identification may be most likely to come into conflictual interactions with environmentalists, and thus develop antagonistic (or negative) ties. With regard to education, in general, much research shows that education tends to be positively correlated with social movement phenomena. In this particular case, we speculate that education is correlated with occupational position, and it is likely that supervisors and managers are more likely to develop organizational roles in the countermovement, and to be more likely to come into contact with environmentalists through things like planning round tables, and perhaps blockades, and similar events. So these factors provide foci (Feld 1982) for such ties to develop.

We speculate that a further reason for this somewhat surprising finding (regarding contending movement ties) is that relatively few members of the environmental movement resided in Port Alberni. As a result, there are few opportunities for counter pressures to develop within the community. As a consequence, the dominant ideology of the community is pro-traditional forestry (pro clear-cutting and industrial forestry) and anti-environmentalism. Perhaps in a more heterogeneous community, the effect for ties to environmentalists would be different—as specified in the original hypothesis (e.g., there would be a greater number of strong contending movement ties).

Finally, we will briefly say a few words for the null ties to people in environmental organization effect for the general public sample. We are confident that the vast majority of members of the countermovement sample are in fact members of the countermovement. However, the general public sample includes some people who are “neutral,” some who are members of the countermovement, and some (albeit likely a smaller number) who are members of the environmental movement. Furthermore, as Table 3 showed, the average member of the general public is less active than is the average member of the countermovement in countermovement activities. Thus, the average member of the general public (because of their lower level of involvement in the countermovement) is less likely to experience antagonistic and conflictual relations with environmentalists. Furthermore, the segments of the general public sample who are neutral, or who in fact are environmentalists, will have different types of relationships with environmentalists. We believe that these factors provide an explanation for this finding.

It might be noted that examining these dynamics are potentially relevant for understanding a number of contemporary movements and controversies such as mobilizations against oil and gas pipelines, climate change skepticism and denial (Dunlap and McCright 2015), and the fight over the survival of the coal industry (Lewin 2019). The approach taken here could be applied to these movements.

Future Research

We undertook this research with the assumption that network processes would operate in a countermovement in a similar fashion to what past researchers have observed for social movements (Diani 1995; McAdam 1986; Tindall 2002). Regarding within-movement ties, this assumption was supported.

Beyond examining within-movement ties, relatively little systematic quantitative research has been conducted to date on either social movements or countermovements regarding the effects of contending movement ties in personal networks. The present study provides a contribution by systematically examining such ties. While some interesting and novel findings are reported here, we would call on researchers to examine further aspects of social networks among countermovement participants.

While we have examined within-movement ties, and contending movement ties, there may be other relevant aspects of people’s personal networks that might be important, including ties to friends or acquaintances (or neighbors, or coworkers) who might support/oppose a particular movement or organizations (e.g., churches, clubs, educational institutions) that might support/oppose a given movement. Examining these, in addition to within-movement, and contending movement ties would paint a more detailed picture of the relationship between social networks and countermovement participation.

Furthermore, we started with the simplifying assumption that network processes would be similar in countermovements, to social movements (and values would be mirror imaged). However, it may be the case that particular countermovements may differ depending upon their substantive content. While the case studied here might be relatively unique (the countermovement is embedded in a relatively homogeneous community that is largely separated from the community the social movement is embedded within), this particular case has similar characteristics to movement-countermovements dynamics around oil and gas, fracking, and coal (perhaps also mining, and dam building) in North America. We might expect countermovements involving these types of natural resource controversies to be similar. However, other types of countermovement may be different because movement participants are embedded in more heterogeneous community networks or perhaps because countermovement participants are more geographically scattered. The affordances of information and communication technologies (ICTs) might affect the development of countermovement social network characteristics and might be particularly important in cases where participants are more geographically dispersed (e.g., recent media reports about members of the alt-right movement). These are but a few topics for future research.

Future research should focus on network processes over time. Key issues include stage of development of the social movement/countermovement, stage in the cycle of protest, and the history of movement-countermovement dynamics. Furthermore, balance theory (e.g., Lerner 2016; Marineau, Labianca, and Kane 2016) and cognitive dissonance may be of some value in understanding the effects of cross embedded ties. Also, it should be added that this research focused entirely on nonvirtual social networks. There are many interesting potential avenues of research on the topics explored in this paper in the context of virtual networks (see Tindall and Groenewegen 2014, Tindall 2015 , Earl et al. 2015).

Finally (as argued previously by McAdam and Paulsen 1993, and Tindall 2007), we would like to emphasize the need for more theoretical and empirical work on the mechanisms that underlie network effects. In the present study, and consistent with theories of group relations, our post hoc analysis of contending movement tie data implies that intense conflict and dislike leads to reinforced commitment. This suggests yet another possible network mechanism. We will end with the following conjecture: the intensity of disliking of alters (associated with contending movement ties) should vary directly with the stage in the cycle of protest, and the relative success of the movement versus the countermovement.