Introduction
It is a Friday morning in May 2023. In a coordinated act, climate activists from the group Last Generation are blocking 12 major roads in Berlin, causing massive traffic jams during rush hour. Wearing orange high-visibility vests, some activists glue themselves to the pavement and hold up banners calling for stronger climate policies. Some drivers react emotionally and decide to take matters into their own hands: Some pull protesters off the road by force, others kick them. After laboriously detaching the last activist from the pavement, the police announce a few hours later that all blockades have been cleared (Barkey Reference Barkey2023; Tagesschau 2023).
This protest is an example of mass obstruction tactics often used by the radical wing of the “new climate movement” (Buzogány and Scherhaufer Reference Buzogány, Scherhaufer, Jörgens, Knill and Steinebach2023), which in addition to Last Generation in Germany includes other organizations such as Extinction Rebellion and Just Stop Oil. These groups use various forms of nonviolent illegal protest such as sit-ins and blockades to disrupt the daily routines of large numbers of people. These actions put activists in a strategic “dilemma” (Feinberg, Willer, and Kovacheff Reference Feinberg, Willer and Kovacheff2020). On the one hand, mass obstruction is highly publicized and draws attention to the protesters’ demands. On the other hand, it increases activists’ exposure to state and vigilante repression. Whether these actions are worth the risk for activists and ultimately help them attain their objectives is the focus of a growing debate in the sociological and political science literature, as well as among the general public.
The impact of mass obstruction on protesters’ ability to achieve their goals likely hinges on how this strategy influences public opinion. Mass obstruction risks creating resentment against those protesting, their demands, and their allies. In this sense, climate activism offers a recent, high-profile example of a recurring strategic problem faced by social movements of all kinds. Although protesters could theoretically achieve their goals by exerting direct pressure on the authorities, swaying public opinion in their favor and building a broad coalition may be more effective, especially since climate activists are demanding large and multifaceted political and societal changes (see Feinberg, Willer, and Kovacheff Reference Feinberg, Willer and Kovacheff2020). A long tradition of research has shown that nonviolent protests generate more public support than violent activism (e.g., Simpson, Willer, and Feinberg Reference Simpson, Willer and Feinberg2018). Yet, when it comes to different types of nonviolent resistance, studies have produced conflicting results about the effects of obstructive tactics such as occupations and blockades (e.g., Bugden Reference Bugden2020; Huff and Kruszewska Reference Huff and Kruszewska2016). This line of research faces two challenges. First, when comparing different types of nonviolent protest, the net effect of obstruction is often confounded with other varying dimensions. Second, studies neglect the role of affected bystanders, who play a key role in the dynamics of these protests, as illustrated in the example above. Indeed, previous studies have shown that protesters’ interactions with vigilantes influence the impact of social movements (Wasow Reference Wasow2020).
We contribute to the literature by examining the public reactions to three types of protest: (1) non-obstructive protest, (2) obstructive protest, and (3) obstructive protest with vigilante intervention. Using a preregistered survey experiment, we carefully manipulated a report of a climate protest to capture the net effect of obstruction and the moderating effect of vigilante intervention on support for the protesters’ actions and goals. Building on theories of radical flank effects, we also tested for spillover effects on support for other climate activists in general, climate groups, and political parties. Our results show that the effects of mass obstruction tactics are mixed: While people judge these actions negatively, support for the protesters’ demands remains remarkably unaffected. Vigilante repression does not moderate reactions to obstructive protests.
The research note is organized as follows. First, we review the literature on social movement tactics and public support. Second, we highlight the particularities of mass traffic obstruction in the context of the new climate movement. Third, we present our hypotheses about the effects of mass traffic obstruction on public support in interaction with vigilante repression. Fourth, we present our research design, and fifth, we discuss our findings and their implications for future research.
Social movement tactics and public support
Social movement activists can use various protest tactics to push for political and cultural change. Research has shown that the degree of disruptiveness of these tactics influences social movement outcomes—along with other dimensions of tactical repertoires such as novelty, variety, size, and cultural resonance (Taylor and Van Dyke Reference Taylor, Van Dyke, Snow, Soule and Kriesi2004). While disruptive forms of mobilization are effective in raising awareness and putting direct pressure on authorities, they can also backfire by limiting the movement’s ability to build a broad base of popular support that may be critical to the movement’s long-term success (Feinberg, Willer, and Kovacheff Reference Feinberg, Willer and Kovacheff2020; for a broader discussion on social movement dilemmas, see also Jasper Reference Jasper2004).
Scholars interested in the relationship between tactical repertoires and public opinion generally agree that by choosing violent over nonviolent methods, social movements damage their reputations and reduce support for their demands. Chenoweth and Stephan (Reference Chenoweth and Stephan2011) suggest that violent resistance campaigns are perceived as less legitimate and therefore tend to be less successful in recruiting new members. Similarly, Simpson et al. (Reference Simpson, Willer and Feinberg2018) argue that the effect of violence on public support operates through identification: Groups that choose violent tactics are perceived as less reasonable, which makes them less relatable and generally more distant from people’s self-image (Simpson, Willer, and Feinberg Reference Simpson, Willer and Feinberg2018; see also Tilly Reference Tilly1995; Wouters Reference Wouters2019). While the exact mechanisms connecting tactical repertoires and public support remain debated, the association between these two concepts has been consistently demonstrated in experimental settings. Reports of challengers using violent actions such as terrorism (Arves, Cunningham, and McCulloch Reference Arves, Gallagher Cunningham and McCulloch2019), bombings (Huff and Kruszewska Reference Huff and Kruszewska2016), and window smashing (Wouters Reference Wouters2019) lead to lower support for the activists compared to reports of challengers relying on nonviolent forms of mobilization.
Beyond the comparison between violent and nonviolent actions, studies contrasting the effects of more or less confrontational forms of nonviolent mobilization have come to conflicting conclusions. Nonviolent mobilization encompasses a wide range of actions, from advocacy to acts of civil disobedience (Brown Reference Brown, Sowers, VanDeveer and Weinthal2023). In a laboratory study, Arves et al. (Reference Arves, Gallagher Cunningham and McCulloch2019) found that participants who were exposed to news stories about rebel groups using demonstrations or hunger strikes (as opposed to participants who were not informed about the groups’ tactics) had more favorable opinions of the groups. Blockades and sit-ins, in contrast, had no significant effect. In their survey experiment, Huff and Kruszewska (Reference Huff and Kruszewska2016) found that individuals who were informed about social movement organizations using occupations—rather than demonstrations—as a strategy did not exhibit significantly higher or lower support for the government initiating negotiations with these organizations. Focusing on climate activism in the US, Bugden (Reference Bugden2020) identified a positive effect of peaceful protest and civil disobedience (compared to a control group) on support for the movement; this effect, however, is solely driven by Democrats and Independents who acknowledge human-caused climate change. The survey experiment by Gonzatti et al. (Reference Gonzatti, Hunger and Hutter2023), conducted in the context of actions by the group Last Generation in Germany, comes to contrasting findings: Compared to peaceful demonstrations, street blockades and the damaging of art pieces lead to lower support for the protest and lower perception of legitimacy. Overall, based on these experimental findings, it remains unclear how more radical nonviolent actions such as occupations and blockades affect public support for protest groups.
These results are difficult to interpret because moderate and confrontational nonviolent actions differ along several dimensions. For example, a peaceful demonstration and the occupation of government offices differ in terms of legality, number of participants, location, and profile of the bystanders. Identifying which of these differences drives the effect on public support is challenging. In this paper, we are particularly interested in mass traffic obstruction, a tactic commonly used by the new climate movement (Buzogány and Scherhaufer Reference Buzogány, Scherhaufer, Jörgens, Knill and Steinebach2023). In what follows, we theorize and empirically test the independent effects of this tactic in a controlled setting, holding other factors constant. We also examine how these effects are moderated by vigilantism, a form of intervention that may significantly alter the dynamics of these actions.
Mass traffic obstruction, vigilantism, and public support
Since the 2010s, the new climate movement has emerged as one of the most influential civil society forces in Europe and North America (Buzogány and Scherhaufer Reference Buzogány, Scherhaufer, Jörgens, Knill and Steinebach2023). Being a large movement, its membership and forms of activism are heterogeneous. Among its members, a handful of groups, including Extinction Rebellion, Just Stop Oil, and Last Generation, employ distinctly confrontational protest strategies that are “intended to physically disrupt people, institutions and/or processes of capital accumulation but that stay short of sabotage” (Berglund Reference Berglund2023, 2).
The radical wing of the new climate movement draws from a broad repertoire of civil disobedience, but one strategy in particular has captured media attention and sparked intense political debate: mass traffic obstruction (MTO). This tactic consists of blocking public transportation infrastructure, such as train tracks, airport runways, and, most often, roads, in order to paralyze traffic on a large scale. MTO intends to cause widespread traffic disruption by targeting high-traffic transportation hubs. These blockades are nonviolent in the sense that the activists block vehicles with their physical presence alone. However, they are also illegal in almost all circumstances and are therefore considered acts of civil disobedience.
In addition to exerting direct pressure on the authorities, the activists’ underlying rationale appears to be to maximize the publicity of the protest to initiate a process of “agenda seeding” (Wasow Reference Wasow2020). In other words, protesters seek to influence the news agenda in favor of their chosen issue by staging confrontational and highly newsworthy protest events. Indeed, given the limited number of activists involved, all of these groups have achieved impressive success in this regard. This is largely due to large-scale disruption but is also aided by certain distinctive and memorable features of their actions. For example, activists from Last Generation have routinely taken to affix themselves to the pavement with superglue, earning them the nickname “Klimakleber” (“climate gluers”) in Germany. As a result, the group has achieved extraordinary visibility, regularly making headlines, not only as an activist organization but as a protest phenomenon in its own right.
We suggest that MTO makes it particularly difficult to stage protests in a way that would help activists gain public support. MTO is deliberately designed to disrupt the daily lives of a large number of citizens, who are more or less arbitrarily affected by the blockade. In the images presented to a wider audience, the affected citizens appear as hostages of the conflict between the protesters and their targets. It becomes very difficult to convey that the protesters’ demands are actually directed at decision-makers who are far removed from the turmoil in the streets. As a result, even a public that is supportive of the protesters’ demands may sympathize more with the affected people at the protest site than with the activists. Moreover, this loss of sympathy opens up opportunities for counter-framing by political opponents (see Benford and Snow Reference Benford and Snow2000).
Creating public resentment through the use of MTO could harm the activists’ cause in several ways. Negative images of the protesters and their actions could affect support for their demands and for other actors who are generally seen as sympathetic. In addition, public resentment creates incentives for the government not to meet the protesters’ demands.
Finally, this form of protest tends to provoke angry and sometimes violent reactions from those affected by the traffic obstruction. The occurrence of vigilantism, i.e., “the extralegal prevention, investigation, or punishment of crimes” (Bateson Reference Bateson2021, 925), must be considered as an additional, relevant factor in the respective staging processes. Somewhat cynically, being the target of vigilante violence may actually help activists achieve their strategic goals. Research on the US civil rights movement found that “when nonviolent protest is met with state or vigilante repression, activists may be injured or even killed, but the resulting media coverage is predicted to help the larger movement and undermine the dominant group-aligned coalition” (Wasow Reference Wasow2020, 643). There is also evidence that violent repression, especially when it is widely publicized and perceived as illegitimate, has positive mobilizing effects (Opp and Roehl Reference Opp and Roehl1990).
Hypotheses
In short, there is no doubt that MTO is effective in attracting attention, but its impact on public opinion remains unclear. Our study aims to fill this research gap by examining how climate activists’ tactical choices, in interaction with bystander reactions, affect public support. Specifically, we are interested in comparing the effects of three types of protest situations in the context of the new climate movement:
T1: an unannounced, non-obstructive demonstration;
T2: an unannounced mass traffic obstruction;
T3: an unannounced mass traffic obstruction with vigilante violence against the protesters.
Compared to an unannounced, non-obstructive demonstration, we expect public support for the protesters and their demands to be negatively affected by the use of MTO. However, when protesters face vigilante aggression, we expect this effect to be attenuated. We formulate the following directional hypotheses:
H1: The type of protest situation influences support for the actions of the protesters.
H1a: Support is higher in T1 than in T2.
H1b: Support is higher in T1 than in T3.
H1c: Support is higher in T3 than in T2.
H2: The type of protest situation influences support for the demands of the protesters.
H2a: Support is higher in T1 than in T2.
H2b: Support is higher in T1 than in T3.
H2c: Support is higher in T3 than in T2.
We also consider that positive or negative reactions may spill over to other groups. Given the heterogeneity of tactics within the new climate movement, moderate groups may benefit (or suffer) from positive (or negative) radical flank effects (Haines Reference Haines, Snow, ella Porta, Klandermans and McAdam2013; Simpson, Willer, and Feinberg Reference Simpson, Willer and Feinberg2022). This is due to a contrast mechanism whereby groups with similar goals but a less confrontational protest strategy may appear more reasonable in comparison. However, depending on the circumstances, they may also be judged guilty by association. This is likely to affect groups in relative alignment (e.g., other environmental groups, green parties) as well as groups in relative opposition (e.g., conservative and radical right-wing parties) in opposite ways. Since the evidence on the direction of radical flank effects in this context is fragmentary, we formulate the following non-directional hypotheses:
H3: The type of protest situation influences feelings toward climate activists in general, activists of Last Generation, and activists of Fridays for Future.
H4: The type of protest situation influences the propensity to vote for major parties represented in the Bundestag.
Research design
In June 2023, we conducted a preregistered survey experiment to examine the effects of mass obstruction and vigilantism on individuals’ perceptions and attitudes toward climate activists, environmental policy, and political parties.Footnote 1 The experiment used a three-group, between-subjects design with participants blinded to their assigned treatment group.
Sample
We recruited a convenience sample online through social media and with a list of email addresses collected over the years by our research unit. Participants had to be residents of Germany and at least 16 years old. After completing the survey anonymously, participants could choose to register for a EUR 10 voucher lottery. Our goal was to obtain a net sample size of 1,550 participants.Footnote 2 We expected approximately 10% missing values due to item non-response and therefore stopped data collection after obtaining 1,700 completed interviews. The average age of respondents was 42 and 67% were male (for a full description of the sample in terms of gender, age, education, and vote choice, see the online supplementary materials, hereafter OSM, Tables S1–S4).Footnote 3
Structure of the experiment
The experiment started with a screening process to ascertain eligibility criteria, wherein respondents were required to provide their age and state of residence. Only respondents aged 16 or older and residing in Germany were permitted to participate in the subsequent survey. Following the screening, participants progressed through a sequence of six steps (see OSM, Figure S1).
The first phase involved eliciting responses to a series of background questions covering political interest, satisfaction with democracy, voting intentions, attitudes toward climate policy, party identification, and car use. The purpose of this set of questions was to establish a baseline understanding of respondents’ predispositions and beliefs. Second, respondents were presented with an introductory page inviting them to read a fictive, but realistic news report about a protest. Participants were instructed to read the report carefully, imagining it to be a real-life event described in a daily newspaper. Third, participants underwent a random assignment process in which they were evenly divided into three different groups. Each group was directed to one of three versions of a news story about a climate protest near the Ministry of Transport in Berlin, corresponding to the three conditions defined in our hypotheses above. Participants could only move to the next page after 25 seconds. In our protest scenarios, we took advantage of the actual location of the ministry: In the first group (our control setting), protesters demonstrated in a park right next to the ministry buildings; in the second and third groups, they blocked the adjacent street. Vigilante repression occurred in the third group, but not in the first and second. The questionnaire (available in the preregistration) presents the full wording of the experimental stimuli. Fourth, following the exposure, participants were surveyed on several outcome measures. Participants were asked to indicate (a) their support for the actions and (b) for the demands of the protesters, (c) their general feelings toward climate groups and climate activists, and (d) their likelihood to vote for the major political parties in the German parliament (see the questionnaire in the preregistration). Fifth, we assessed participants’ ability to accurately recall two aspects of the experimental setting: whether the climate activists blocked a street and whether a driver tried to separate a protester from his companions.Footnote 4 Sixth, at the end of the survey, we collected sociodemographic information about the participants, such as their gender, occupation, education, living environment, household composition, children, income, and party membership.
Results
To test our hypotheses, we estimated one-way ANOVAs and used Tukey’s HSD testFootnote 5 to compare group means across all possible pairs of groups. We hypothesized that support for the protesters’ actions (H1) and demands (H2) would be highest when respondents are exposed to a non-obstructive demonstration and lowest when they are exposed to a mass traffic obstruction without vigilantism, with the mass traffic obstruction with vigilantism scenario falling in between. Figure 1 shows the results. Respondents who read the first scenario of a protest without a blockade rated the actions with a mean value of 8.0 on an 11-point scale ranging from 1 “do not support it at all” to 11 “fully support it.” Respondents in the second and third groups rated the actions significantly lower with 5.9 and 5.6, respectively. These results support hypotheses H1a and H1b: The protests in which a street was blocked led to significantly less support for the actions of the protesters – but there is no significant effect of vigilantism on the evaluation of the actions (H1c). In addition, there are no significant differences in the evaluation of the protesters’ demands between the three scenarios (H2). In all three scenarios, respondents support the demands with similar values. The evaluation of the demands is independent of the specific protest behavior.
We also hypothesized that the type of protest situation would influence feelings toward climate activists in general, activists of the group Last Generation, and activists of Fridays for Future (H3), as well as the propensity to vote for the major parties represented in the Bundestag (H4). These hypotheses must be rejected. The differences between the three groups are negligible and do not reach statistical significance (see OSM, Figures S4–S12). Neither the evaluations of different climate activists nor the propensity to vote for one of the major parties in Germany are affected by the protest scenario.
We conducted two additional analyses. First, we excluded respondents who failed the manipulation checks. The results remained constant in each case (see OSM, Figures S13–S22). Second, we conducted subgroup analyses to control for potential heterogeneity within the results. We contrasted the results for the full sample with those for eight subsamples: younger respondents (under 40), older respondents (40 and older), female respondents, male respondents, only Green party voters, all respondents except Green party voters, (center) right party voters, and regular car users.Footnote 6 As expected, we found differences in the baseline levels of our outcome measures, but generally no differences in the effects of the treatment (see OSM, Figures S23–S42). One interesting exception that deserves further investigation is that Green voters exposed to the vigilante violence scenario rated the activists’ actions lower than when they were exposed to the blockade alone. This suggests that Green voters—who are natural allies of the activists—may be particularly affected by the backlash created by mass obstruction tactics. However, this does not affect their support for the protesters’ demands.
Discussion and conclusions
This study examined the effects of mass traffic obstruction, a controversial tactic used by parts of the new climate movement. This type of protest has been costly for the climate activists, but it has also brought them extraordinary media attention. In an online experiment, we examined how the effects of this tactic on public support for the protesters varied in two different settings—with or without vigilante violence. We found that mass traffic obstruction is viewed less favorably than a similar non-obstructive protest. However, contrary to popular belief, mass obstruction did not significantly affect people’s positions on the substantive issue of climate change. Moreover, we found no spillover effects on opponent and allied groups. Finally, vigilantism against the protesters had no discernible effect on public opinion in our full sample.
Our study has allowed us to test the causal effect of mass obstruction tactics, in interaction with bystander reactions, on public opinion. However, our experimental design comes with certain limitations. Our treatment may not have been strong enough to overwrite prior mental images created by extensive media coverage. Subtle differences in protest reports may not have created sufficiently distinct mental images to significantly influence opinion. Moreover, for ethical reasons, we could not expose participants to more intense (and realistic) forms of vigilante repression. Our results might thus underestimate its effect.
Based on these findings, what should we conclude about the impact of mass obstruction on activists’ ability to achieve their goals? On the one hand, mass obstruction is unlikely to help activists build a larger coalition. Challenger groups using this strategy have little chance of persuading undecided people to support their cause. On the other hand, these radical actions are unlikely to change the minds of those who already support the protesters’ demands. For activists, this means that mass obstruction tactics might be “worth the risk” because they put pressure on authorities and attract media attention without causing people to defect to the other camp or damaging the reputation of allied groups (see Feinberg, Willer, and Kovacheff Reference Feinberg, Willer and Kovacheff2020). The consequences of the new climate movement’s disruptive tactics may therefore be more limited than expected. As research on the animal rights movement has shown (Jasper and Poulsen Reference Jasper and Poulsen1993), this may be especially true at the beginning of a campaign, before the legal costs of mass obstruction begin to mount and other non-state actors build up a strong counter-movement.
Future research should continue to explore how climate activists weigh the pressures of internal and external audiences when choosing tactics like mass traffic obstruction (Jasper Reference Jasper2004) and how patterns of responses to different tactics within populations are affected by external shocks (Cristancho, Uba, and Zamponi Reference Cristancho, Uba and Zamponi2019), particularly exposure to the impacts of climate change.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1755773925000098.
Data availability
The data and analysis code used in this study are publicly available via the Open Science Framework (OSF) at https://doi.org/10.17605/OSF.IO/4ZAYB.
Acknowledgements
We thank Thorsten Faas, Chiara Franke, Teresa Haußmann, and Katrin Schmitz for their support during the data collection and preparation of the manuscript.
Funding statement
The authors do not report funding for this project.
Competing interests
The authors do not report any conflict of interest.
Open access
