Demobilising far-right demonstration campaigns: Coercive counter-mobilisation, state social control, and the demobilisation of the Hess Gedenkmarsch campaign

ABSTRACT Studies of social movements have often focused on mobilisation and campaigning; by comparison, demobilisation has received little attention. This article adds to the body of literature on demobilisation by examining one case of demonstration campaign demobilisation. The ‘Hess Gedenkmarsch’ campaign in Germany, initiated in the late 1980s and demobilised by the mid-1990s, is not only a case of a causal mechanism of demobilisation, but also particularly important within far-right social movement activity: it was the vanguard campaign in a emergent pattern of ‘demonstration politics’ by far-right groups in Germany. The case exhibits a process whereby anti-far-right activists effectively engaged in a sort of kamikaze counter-mobilisation, seeking to shut down far-right events; this, in turn, spurred state authorities to act, imposing coercive measures that demobilised the far-right campaign. This case illustrates a causal mechanism of negative demobilisation that can be observed in other demonstration campaigns, and is particularly relevant to other cases of far-right activism.

and introducing new issues, frames, and forms of action that can be used in subsequent campaigns.' The process of campaign demobilisation is thus an important part of the wider examination of SMO and movement lifecycles.
Amongst the most common campaign forms are demonstration campaigns. That whole campaigns consist of series of demonstrations is not surprising: demonstrations serve several important purposes for SMOs: raising awareness, attracting new members and keeping existing members connected, promoting group solidarity, facilitating networking. Beyond such instrumental ends, demonstrations provide meaningful references to past events, preceding mobilisations within social movements, and can fulfil other symbolic objectives.
These various ends arguably obtain a particularly high importance among far-right SMOs. Regular, big events stand out like mile-markers in the far-right activists' calendar, moments to assemble, make or renew connections, and present the far-right movement in public. As a result, there is typically not much internal pressure to abandon demonstration campaigns. Neither should nostalgia be discounted; referencing previous mass demonstrations by the far right figures in the motivations for many activists. Even today, when on-line communication and movement organisation attracts great public and scholarly attention, off-line, subculture-based mobilisation, grounded in events like demonstrations and concerts, remains particularly important for far-right groups.
Negative demobilisation of demonstration campaigns has two clear manifestations. First, a campaign ceases its initiation of demonstrations. If organisers stop their interventions (Davenport, 2015, p. 21), the campaign has demobilised. Second, demonstration campaigns live and die by the number of individuals mobilised for the event; in some respects, it does just come down to numbers (Biggs, 2018;Denardo, 1985). So when a demonstration campaign experiences decreased participation, it undergoes a degree of negative demobilisation -all the more so when the decrease is large and sustained over successive demonstrations.
Below, the paper examines one case of negative demobilisation: the so-called 'Hess Gedenkmarsch' (i.e., Hess Memorial March) campaign in Germany, initiated in the late 1980s and demobilised by the mid-1990s. 3 The case represents a process that can be observed in other demonstration campaigns, but it is also particularly important: it was the vanguard campaign in a broader trend of 'demonstration politics' by far-right SMOs in Germany (Virchow, 2007(Virchow, , 2013a(Virchow, , 2013b. It developed into one of the largest far-right gatherings in Germany, and acquired transnational significance with the participation of activists from throughout Europe. Before delving into this case, the next section theorises a causal mechanism of coercive counter-mobilisation triggering state social control

A causal mechanism of negative demobilisation by coercive counter-mobilisation and state social control
One pattern of negative demobilisation of far-right demonstration campaigns is the sequence of violent (non-state) counter-mobilisation followed by state social control 4 (i.e., repression). This pattern is observable in several cases -violent countermobilisation intent on shutting down far-right demonstrations is the most noted modus operandi of militant anti-fascist movements -yet the causal process has scarcely been unpacked in detail.
This section is a corrective; it theorises a mechanism that explains how a far-right demonstration campaign met by violent counter-mobilisation produces negative demobilisation ( Figure 1). In unfolding the parts of this mechanism, I identify the observable manifestations that reveal its presence. This relies on the detection of two links of sequence evidence -namely, that violent counter-mobilisation occurred before state social control, which occurred before negative demobilisation of the demonstration campaign -and several pieces of account evidence, that is, contemporary material that confirms or suggests the operation of the mechanism (Beach & Pedersen, 2013, pp. 99-100). The following sections evaluate case observations on the basis of the likelihood ratio, which is equal to the probability of finding predicted evidence if a hypothesis is false (i.e., evidence 'uniqueness' 5 ) divided by the probability of finding predicted evidence if a hypothesis is true (i.e., evidence 'certainty' 6 ). 7 The mechanism applies to localised demonstration campaigns and is expected to function in stable states with open, plural public spheres, and where state authorities have the capacity to impose social control on demonstrations. 8 The former condition is necessary for opposed, violent demonstration mobilisations to occur (the first part of the mechanism); the latter, necessary for state social control to result ultimately in negative demobilisation (the second part of the mechanism). Cases representing this negative demobilisation process will have the characteristic one-two punch: violence between antagonistic demonstrations followed by state social control. To be sure, 'mechanisms have different empirical manifestations in different cases' (Beach & Pedersen, 2013, p. 73), but this simple but significant causal sequence stands on a level of abstraction that permits generalizability to myriad other cases.

Cause: far-right demonstration campaign
This specific demobilisation starts with the mobilisation of a far-right demonstration campaign. There are two manner of stimuli within this cause. First, large, far-right demonstration campaigns may attract the attentions of violent counter-mobilisation by virtue of their size: the larger a demonstration, the greater its potential impact, generally (Biggs, 2018;Denardo, 1985). Big demonstration campaigns, therefore, are big targets. Second, campaigns may invite counter-mobilisation by the particular far-right sympathies given form by demonstrations: stark glorification of Nazism in modern Germany, for example, is highly provocative; sheer egregiousness alone, irrespective of campaigns size, may invite the attention of anti-fascist groups. These conceptions have distinct manifestations in the case of the Hess Gedenkmarsch (Table 1).
C 1 (part I of causal mechanism): Coercive counter-mobilisation triggers state social control Whether because of its size or egregiousness, far-right demonstration campaigns attract the attention of oppositional activists. Campaigns fight for attention and effect; against the multiplicity of coincident actors, yes, but most intensely against forces that explicitly, doggedly resist their efforts. Mobilisation often provokes counter-mobilisation, the reactive process of social forces responding to threats, organising resistance against changing or changed opportunities. In all its various forms it is an attempt to raise the costs of an initiating mobilisation's campaign. It is an exercise of private (i.e., non-state) social control. This is the first link in part one of the causal mechanism -and violence, coercive action is an essential characteristic of it. But it is crucial to note that violent counter-mobilisation is not so much a product of any particular far-right campaign as it is a fundamental tenet of some groups that respond to far-right activism.
Far-right activism long ago spawned countermovements, such as 'Antifa,' 11 which persistently challenge their mobilisation. Mutual antagonism between oppositional movements gives rise to recurrent triadic interaction between mobilisation, countermobilisation, and state authorities.
Some countermovements 12 and counter-mobilisations against the far right are nonviolent. Yet others are, if not consistently violent, convinced that violence can be an effective means of demobilising the far right; this is often termed 'militant anti-fascism' (e.g., Bray, 2013;Copsey, 2000). Militant anti-fascism stems from a conviction, essentially Social actors that fit within one or more of Minkenberg's four variants of radical right forces 10 Demonstrations honouring a famous Nazi and close acolyte of Adolf Hitler; demonstrations organised by avowedly neo-Nazi activists and peopled by neo-Nazi movement members Large Amassing more than 1000 participants at peak mobilisation of the campaign The number of participants reported in Hess Gedenkmarsch events by police and news outlet accounts.
based on two observations: at the micro-level, far-right activism can be deadly even when small and isolated. (Numerous cases of far-right murders and recent scholarship on farright terrorism lend credence to this contention). At the meso-level, far-right organisation is still a threat even when small. Mussolini's Blackshirts grew from around 100 men in 1919 to a quarter of a million in less than two years; the Nazi party had only fifty-four members when Hitler first attended one of its meetings in 1919 (Bray, 2013, p. 140). These two considerations make some violent counter-mobilisation against the far right an artefact of the countering group's internal culture -not a product of any particular situation. Thus, when a far-right demonstration campaign attracts the attention of militant anti-fascist groups, like Antifa and Autonomen groups, aggressive action is the likeliest of any counter-mobilisation form. The first step in the causal process depicted in Figure 1, therefore, is the far-right demonstration campaign attracting the attention of a militant anti-far-right group(s). One observable manifestation that does not confirm, but makes more plausible this step is whether the demonstration events or the campaign itself was widely reported or reported in sources readily available to militant anti-far-right group(s) (OM1). Of course, this reporting must precede the violent counter-mobilisation for this part of the mechanism to hold. Observation of this reporting has a low likelihood ratio because, while it almost certainly should be observed to confirm the hypothesised part of the causal mechanism, it is hardly unique: reporting only establishes circumstantial grounds for presuming that militant anti-far-right groups are aware of the campaign. What would provide strong confirmation is testimony from anti-far-right activists that the far-right demonstration campaign had come to the attention of their group and the intention to counter-mobilise was decided and planned (OM2). While this observation would have a high likelihood ratio, strongly supporting the hypothesised part of the causal mechanism, it is unlikely to be observed for two reasons: (1) anti-far-right groups are often wary of research and investigation for the risk it can pose to identification and reprisals from far-right activists, so they often constitute 'hidden communities'; (2) the longer research is temporally removed from actual events, the less likely it will be to find reliable witnesses and strong confirmatory evidence. However, it is not an implausible leap from the circumstantial grounds of OM1 to clear evidence that the substance of OM2 (i.e., militant anti-far-right groups were aware and decided to counter-mobilise) did occur, if there are references to the campaign and calls for counter-mobilising actions in militant anti-far-right groups' publications and internal documents (OM3). While the absence of OM2 would obscure some of the sub-processes, OM3 nevertheless provides high likelihood that confirms the hypothesised part of the causal mechanism.
Next, counter-mobilisation results in a significant amount of violent altercations between antagonistic demonstrators. While the distinction between 'a significant amount of violent altercations' and 'a few marginal incidents of violence' is blurred, reporting of violence provides some clarity. Extensive reporting of violent altercations (OM4) spurs the transference of causal energy from the violent events to local state authorities. Observations of this are weighty because violence at a large demonstration event is almost certain to be reported, and because alternative explanations for this evidence, short of unreliable reporting (which decreases when multiple independent sources offer similar accounts), are improbable. Similarly, reporting may include, result in, or coincide with heightened public safety concerns as a result of violent demonstration events (OM5). But this is fickle evidence; many alternative explanations are possible (other coincidental occurrences that elevate safety concerns among local residents), so it is not unique; nor is it certain for it may or may be observable in a case, and because heightened concerns may register among the citizenry and state authorities or among state authorities alone. Nevertheless, what is essential to the hypothesised mechanism is that these concerns register with state authorities, driving them to adopt social control measures on the far-right (and other) demonstrations (OM6). In stable states with open, plural public spheres, and where state authorities have the capacity to impose social control on demonstrations, such measures typically must be justified on legal bases. 13 This initial adoption of measures and justification may or may not be publicised, but where available it would confirm the activity of local authorities. Notification of this decision to the demonstration organisers (OM7) would also provide strong confirmation, but is also unlikely to be observed. However, because far-right organisers tend to be committed to continuing their campaign and because many contexts offer encouraging opportunities for overturning social control measures, 14 a judicial review of social control measures often ensues. The first part of the causal mechanism only functions if the judiciary upholds the social control measure(s)otherwise, demobilisation may not occur, or only by some other process. Record or accounts of a judicial decision confirms (at least implicitly) OM6 and OM7, even if precise details of these manifestations remain obscure.
In sum, these observable manifestations (Table 2) comprise the first part of the causal mechanism. They provide the two essential pieces of sequence evidence -that violent counter-mobilisation occurred before state social control, which in turn occurred before negative demobilisation -and account evidence that confirms the transference of causal energy in the process.
C 2 (part II of causal mechanism): Attempted far-right adaptation meets persistent social control The judicial confirmation of local authorities' social control of far-right demonstrations might lead directly to negative demobilisation -but this is not typical. Far-right SMOs are often so wedded to their strategy of demonstration campaigns -not surprisingly given the aforementioned advantages -that organisers attempt to 'beat the ban' with small tactical adaptations. That is, organisers may attempt to excise the proscribed characteristic(s) of demonstrations and reapply for permission, or to change slightly the Table 2. Observable manifestations (OM) of C 1 (part I of causal mechanism).

No.
Observable manifestation (OM) 1 Events of far-right demonstration campaign reported widely and/or in sources readily available to militant anti-far -right group(s) 2 Testimony from anti-far-right activists confirming that the far-right demonstration campaign had come to the attention of the group and the intention to counter-mobilise was decided and planned 3 References to the far-right demonstration campaign and counter-mobilising actions in militant anti-far-right group(s) publications and internal documents 4 Accounts of significant violent altercations -not merely a few marginal incidents -between far-right demonstrators and counter-mobilising activists 5 References to heightened public safety concerns in reporting of event, in testimony from local residents or in the actions of local state authorities 6 Account of local/presiding state authority adopting social control measures on legally justified bases 7 Account notification of decision to organising far-right group/activist, and decision to appeal to legal adjudication 8 Record of judicial decision confirming local state authority's social control measures date or location of the event, and other such innovations while maintaining large-scale demonstrations as the central tactic. Accounts of such tactical adaptation (OM9) indicate that the first punch of state social control did not result in the campaigns negative demobilisation. Indeed, if social control measures are irresolute, demonstration campaigns may carry on with minor adaptation. Two factors do push the far-right demonstration campaign downward to negative demobilisation: continued counter-mobilisation and continued state social control. Accounts of continued counter-mobilisation (OM10) represent the continued threat to public safety. This threat compels state authorities to respond with further social control measures (OM11). The astute reader will note that this echoes the first part of the causal mechanism, except that local authorities may find implementing further social control easier because of the pattern set previously. Alternatively, it is possible that state authorities act to control (and demobilise) the campaign without persistent countermobilisation -but it would be curious given its failure to do so earlier, before the occurrence of violent counter-mobilisation. In any case, continual imposition of social control on the campaign deprives the far right of satisfactory scope to continue its campaign and hastens negative demobilisation.

Outcome: negative demobilisation
The preceding parts of the causal mechanism produce negative demobilisation in one of two forms. First, the SMO or group of SMOs ceases organising demonstration events. This could be strongly confirmed by account evidence from the SMO stating that no further demonstrations are planned (and better still, explaining why) (OM12). Yet this is highly unlikely. With leaders' pride and movement prestige at stake, open admission of being forced off the street is improbable; and esteeming such evidence would mean taking the partisan explanation of movement activists at face value. It is sufficient to find no accounts of further demonstration events (OM13).
Yet, second, negative demobilisation of the campaign may occur in degree when participation in demonstration events significantly decreases (OM14). Denardo (1985, p. 36) argues, 'the disruptiveness of protests, demonstrations, and uprisings [is] first and foremost a question of numbers.' A significant decrease in demonstration participation carries a corresponding decrease in the benefits SMOs can derive from the event. When participation halves or diminishes even further over successive events, a form of demobilisation is occurring. Empirical observation of this is likely as police are typically still called upon to monitor and superintend events, and some journalistic coverage is to be expected, even of smaller protest events. The evidence would provide a high degree of certainty -diminished participation must be observed to qualify as this sort of negative demobilisation -and a moderately high degree of uniqueness -of course, reporting of numbers is manipulable and subject to ulterior motives, but these liabilities decrease with confidence in the veracity of sources and the addition of independent accounts. Each of these observable manifestations (Table 4) signifies the negative demobilisation of a farright demonstration campaign.

The Hess Gedenkmarsch campaign, 1988-1994
On 17 August 1987, Rudolf Hess, acolyte of Adolf Hitler and last remaining prisoner of the Second World War, committed suicide in his cell in Spandau Prison in Berlin. Far-right activists, led by Michael Kühnen, Berthold Dinter and his Hilfsorganisation für nationale politische Gefangene organisation, and Christian Worch 16 initiated an annual 'Gedenkmarsch' ('memorial march') to his grave in Wunsiedel. This campaign (summarised in Table 5) materialised within a context of swelling far-right activism in West Germany (Lee, 1997, p. 234), where demonstrations were seen as fruitful means of developing the far-right movement (Virchow, 2007, pp. 296, 301). By 1990, with involvement of participants from East Germany, the campaign mobilised well over a thousand far-right activists.

Cause: the emergence of the Hess Gedenkmarsch campaign
Rudolf Hess's death in 1987 triggered some small outpourings of far-right mobilisation; after all, Hess had been the cause of some low-level mobilisation, advocacy, and idolisation before his death (O'Hara & Schlueter, 2002). But 1988 marked the beginning of the campaign: Hess was interred in his family's plot in Wunsiedel in March, and far-right leaders -principally, Berthold Dinter, Michael Kühnen, and Christian Worch -and their SMOs planned a demonstration event for the anniversary of his death. They thought to make of Hess a martyr. 17 Initially, the local council (Landratsamt) of Wunsiedel, the presiding local state authority (the Versammlungsbehörde), sought to ban the demonstration, but the decision was overturned on appeal and the demonstration allowed to take place (O'Hara & Schlueter, 2002, p. 17). 17 August 1988 saw over a hundred neo-Nazis march through Wunsiedel and pay homage at Hess's grave. The following year saw participation nearly double, owing to the participation of the Freiheitliche Deutsche Arbeiterpartei (FAP), that is, the 'Free German Workers' Party' (O'Hara & Schlueter, 2002, p. 17;Virchow, 2013a, p. 175) -though still a fairly inconspicuous demonstration of around 200 participants.
The context surrounding this emergent campaign shifted massively in 1990. When the Berlin Wall fell, as many streamed into West Germany, several far-right leaders from the West moved into the East (Lee, 1997, p. 237;Bundesministerium des Innern, 1991, p. 117). The opening of the East offered up vast, untapped mobilising networks to farright organisers. In January 1990, a group of far-right activists occupied a house in Berlin-Lichtenberg (Weitlingstrasse, 122) to use as a base of operations for mobilising far-right sympathisers in East Germany (Bundesministerium des Innern, 1991, p. 118). Well before the annual demonstration in August 1990 it was clear there would be dramatically increased participation. In addition to being a stridently far-right demonstration campaign, honouring one of the most famous figures of the Nazi regime, 1990 also promised to make the Hess Gedenkmarsch a large event.
C 1 (part I of causal mechanism): Antifa and Autonomen counter-mobilisation, and local authority social control Plans for a far-right demonstration did not go unnoticed. The death of Rudolf Hess in 1987 was a major event. He was the last prisoner of the Second World War, and he died in Berlin in Spandau Prison. Memorialisation of Hess was publicised by East German periodicals -unsurprisingly, since communist authorities were often keen to accentuate the far-right under-currents in capitalist states. The largest daily newspaper in East Germany, Neues Deutschland, reported on the 1989 event (Table 2, OM1) with the headline 'Police protect neo-Nazi march in Oberfranken,' and wrote that there were 500 participants ("Polizei schützte Neonazi-Aufmarsch in Oberfranken, 1989).
Anti-fascist activism was considerable, particularly in Berlin where it was closely intermingled with the 'squatters movement.' The weekly publication of the 'Squatters Newspaper' (HausbesetzerInnen Zeitung) imparts the unmistakeable impression of a movement that is thoroughly aware of even minute current events and of how its ideological opponents organise. There was little ambiguity about dealing with certain opponents, moreover: 'Fascism, racism and anti-Semitism are not political perspectives, they are crimes. One does not discuss with fascists, one forbids the fascists' ("Nummer 1., 1990, p. 10). This quote captures the SMO's typical tone regarding far-right activism.
This particular group, one part of the community of anti-far-right organisations, evinced its awareness of the Hess Gedenkmarsch and declared its intent to countermobilise (Table 2, OM3). Several editions of the SMO publication announced plans to travel to Wunsiedel to confront the far-right event in 1990. 18 Other leftist publications, such as 'Telegraph' ("Antifaschistischer Sommerkalender, 1990), announced plans to oppose the demonstration, too. West German state security services also note that the 'Antifa Telephone' hotline was used to counter-mobilise (Bundesministerium des Innern, 1991, p. 52).
The counter-mobilisation was massive. State security services reported as many as 4000 counter-protesters, including around '800 Autonome and other violent leftextremists' (Bundesministerium des Innern, 1991, p. 52). Reporting from Die Zeit newspaper claims that 2000 counter-demonstrators took part in the peaceful counter-demonstration march that preceded the far-right event (Drieschner, 1990). Police were deployed in large numbers to separate opposing demonstrators, but militant anti-far-right activists fought with far-right demonstrators on several occasions (Table 2, OM4), as well as clashing with riot police (Bundesministerium des Innern, 1991, p. 52). Television reporting from the event captured some of the running battles that took place later in the day (Klocke, 1990). Violence was extensive and posed a significant policing challenge.
Several sources attest to the concerns of Wunsiedel residents (Table 2, OM5): in television interviews locals spoke with exasperation about the disruption caused by the demonstrations (Klocke, 1990). The local branch of the Christian Social Union (Christlich-Soziale Union, CSU) party, which led the local council at the time, collected 600 signatures for a petition to ban future demonstrations ("Wunsiedel Stadt Chronik 1974-1991, 1991. These manifestations of elevated concern and public desire for preventative action do not directly confirm a transfer of causal energy which moved local state authorities to act -but it makes the presumption of transference exceedingly plausible. The city chronicle records that on 25 June 1991, the local council of Wunsiedel banned the Hess Demonstration (Table 2, OM6) (Wunsiedel Stadt Chronik 1974-1991, 1991, applying the powers granted to presiding state authorities under section 15 paragraph 1 of the 'Assembly Law' (Versammlungsgesetz). The section states, The responsible authority may prohibit the meeting or the procession, or make it dependent on certain conditions if, according to the evident circumstances at the time the order is issued, public safety or order is imminently endangered by the meeting or the procession ("Gesetz über Versammlungen und Aufzüge, 1953).
Writing later about the Hess Gedenkmarsch campaign, Peter Pezolt, a senior officer in the Bavarian Police Directorate, explains that the 'massive riots' of 1990 gave the local council grounds to ban the events, since similar incidents were expected; 'the maintenance of public security and order, even from the perspective of proportionality, was only possible by banning the registered assembly' (Pezolt, 2008, p. 3). Here, we have firm confirmation of the activities of entities and, more importantly, their rationale.
Items from the city chronicle testify to the judicial confirmation ( Thus, the state enacted, confirmed, and enforced social control measures. This completed the first part of the causal mechanism, prompting (as the next section addresses) adaptation by the far-right campaign. One curious fact stands out from this first part: judicial decisions justified the ban both on the grounds of the ideological character of the Hess Demonstration and on the credible concerns of violence between opposed demonstrators (Pezolt, 2008, p. 4). However, given that the ideological character of the campaign was consistent from the start -honour Rudolf Hess and (at least implicitly) the Nazi regime -it seems that either the sheer size of the far-right event or (more plausibly, given the available evidence) the presence of aggressive counter-mobilisation was the impetus for state action.

C 2 (part II of causal mechanism): Far-right adaptation met by persistent countermobilisation and social control
Far-right organisers did not yield to this first volley of counter-mobilisation and social control. Indeed, the upwards of 1100 participants in 1990 consisted of several far-right activists from other European countries. In the immediate aftermath of the 1990 event's tumult, far-right leaders were outwardly buoyant. Christain Worch, one of the principal organisers, reportedly said (O'Hara & Schlueter, 2002, p. 18), 'We assume that the left-wing counter-demonstration will not be allowed to take place next year . . . Then a situation will arise in which comrades are no longer predominantly just the young, radical comrades ready to defend themselves will come. The good citizens of the National Camp, the tie-and collar-wearers and many older people who stayed at home for fear of stone-throwing will join.' Given this hope -not unfounded -the original (now banned) event was changed to a protest in Bayreuth, in front of the Administrative Court that upheld the demonstration ban. Far-right lawyer Jürgen Rieger, along with Worch, mobilised at least 1500 activists to this demonstration (Table 3, OM9) (O'Hara & Schlueter, 2002, p. 18;Vierkant, 2015, p. 273;Virchow, 2013b, p. 184). A large police contingent separated these activists from as many as 2500 counter-demonstrators (Table 3,  1992 saw widespread diffusion of state social control measures (Table 3, OM11). Not only was a demonstration ban still in place in Wunsiedel, but attempted registration of demonstrations also met with bans in at least 20 locations in Thuringia and Saxony (O'Hara & Schlueter, 2002, p. 18;Virchow, 2013a, p. 176). A tactical innovation enabled the far right to arrange a demonstration: a far-right activist, recounting the events, explains that the far-right demonstrators drove in separate convoys but, 'armed with auto-telephones' (Table 3, OM9), stayed in constant contact (Virchow, 2013b, p. 176), and so manoeuvred to Rudolstadt, a town in Thuringia, where they demonstrated Table 3. Observable manifestations (OM) of C 2 (part II of causal mechanism).

No.
Observable manifestation (OM) 9 Accounts of far-right adaptation to social control measures (e.g., holding proxy demonstrations, changing date or location) that, in any case, preserves the tactic of large-scale demonstrations 15 10 Accounts of continued counter-mobilisation against the campaign 11 Accounts of similar social control measures adopted by local authorities to deal with adaptations of far-right demonstration campaign Table 4. Observable manifestations (OM) of the outcome. No.
Observable manifestation (OM) 12 Documents, publications, or accounts from/of the organising far-right SMO or its leaders stating that continued social control or counter-mobilisation make more demonstration events impracticable 13 No accounts of further demonstration events 14 Accounts of further demonstration events consistently record significantly decreased participation in demonstration events  (Table 3, OM10, OM11), along with the re-issuance of demonstration bans in several towns in central Germany. An article in Neues Deutschland announced Antifa's intention to prevent the Hess demonstration, wherever it would actually take place (Diesmal in Wiesbaden?, 1993). In reports published later, Antifa activists claim that they got access to the far-right hotline number and learned that the demonstration was planned for Bischofferode (Thuringia) (Antifaschistischen Nachrichten, 1994). Evidently, police were alive to the situation as well because 5000 police officers were deployed to the town (Antifaschistischen Nachrichten, 1994;Klingelschmitt, 1993). Police prevented assembly in Bischofferode, arresting 28 farright activists; they attempted to hinder counter-mobilisation by stopping cars and buses of Antifa activists and confiscating the telephone cards that they were using to coordinate their convoys (Antifaschistischen Nachrichten, 1994) 20 -though Antifa activists underscored their continued intent to disrupt the far-right mobilisation with acts of violence and property damage (Bundesministerium des Innern, 1994, p. 40). The remnants of the far-right activists assembled in Fulda and demonstrated -significantly smaller than in previous years, harried still by state and private agents, but nevertheless persisting.

Outcome: negative demobilisation
The 1994 edition of the Hess Gedenkmarsch was decidedly a debacle and effectively marked the demobilisation of the far-right campaign. Bans were again widespread and police again out in force. Antifa activists counter-mobilised and demonstrated violently (Bundesministerium des Innern, 1995, pp. 43-44). Some far-right activists held small commemorations (never numbering more than a few dozen activists) for Hess -but this nullified most of the advantages that the far-right movement derived from demonstrations in previous years. Still, far-right organisers, seeking further tactical adaptation, thought to hold a demonstration in the neighbouring country of Luxembourg, where the legal context seemed more permissive. Each of the 180 far-right activists who demonstrated in front of the German embassy in Luxembourg (Table 4, OM14) were arrested (O'Hara & Schlueter, 2002, p. 20).
From 1994 onward state social control against the Hess Gedenkmarsch campaign was extensive and effective. No major events were held in commemoration of Rudolf Hess and, at least for a time, this far-right demonstration campaign negatively demobilised. A new manifestation of the Hess Gedenkmarsch campaign materialised in 2001, led by right-wing lawyer Jürgen Rieger, a veteran organiser who was involved in the first campaign. Given this, it is possible to conceptualise 1994-2000 as a campaign in 'abeyance,' 21 but this would present more of a semantic distinction from 'negative demobilisation' than a substantive one; even in so doing, one recognises an end (e.g., Virchow, 2013b, p. 176), albeit temporarily.
The campaign's negative demobilisation (see Figure 2), moreover, came amid wider, committed applications of state social control. Spikes of incidents of far-right violence, including several demonstrations-cum-pogroms (e.g., in Hoyerswerda in September 1991 andin Rostock in August 1992) and arsonist attacks (as in Mölln and Solingen), provoked considerable public pressure on the state to act. Events like commemorations of Hess, symbolic of Germany's far-right scene, were suppressed. Several organisational bans against far-right groups were implemented, including on groups that participated in the Hess campaign, such as the Wiking Jugend (banned in November 1994) and on the National Liste and FAP (banned in February 1995).
The manner in which a new Hess Gedenkmarsch campaign materialised in 2001 provides a final piece of confirmation of the causal mechanism of violent countermobilisation prompting state social control prompting negative demobilisation. Several rulings in 2000 and early 2001 by Germany's Federal Constitutional Court (Bundesverfassungsgericht) nullified assembly bans against far-right events ordered by local authorities (e.g., Bundesverfassungsgericht, 2001). Oftentimes, the court's decisions asserted that there was not enough evidence of a danger to public security to justify a ban; in early 2001 the Bavarian High Administrative Court upheld Jürgen Rieger's complaint, allowing his Hess Gedenkmarsch event to assemble in Wunsiedel explicitly because there was not a sufficient threat from anti-fascist counter-mobilisation (Pezolt, 2006, p. 259;O'Hara & Schlueter, 2002, p. 23).

Conclusion
What is important about this case? What does it tell us about causal processes of demobilisation? The Hess Gedenkmarsch campaign shows that social control, repressive measures can bring about negative demobilisation. While this finding is not in itself surprising, the process surrounding it is. Anti-far-right activists effectively engaged in a sort of kamikaze counter-mobilisation, seeking to shut down far-right events; that the demonstration bans imposed by local authorities applied equally to Antifa and the farright demonstrators was of no matter.
Before 1991, the state had the means and opportunity to press for the demobilisation of the Hess Gedenkmarsch, but evidently the motive to apply its means, at least in the view of judicial authorities, was insufficient until the appearance of violent countermobilisation. The insufficiency for assembly bans until 1991 centred around the threat to 'public safety and order,' but state authorities might well have accepted justification on these grounds: state security authorities considered many of the far-right activists attending the demonstrations violent (gewaltbereit); forbidden signs and symbols, such as the so-called 'Kühnen salute,' were used, yet judicial authorities would not confirm a ban. In other words, violent counter-mobilisation, for better or worse, can affect negative demobilisation: by triggering state social control, as in the Hess Gedenkmarsch case.
Counter-mobilisation and state social control deprived the Hess Gedenkmarsch of the benefits that far-right organisers sought. Part of the symbolic raison d'être was withdrawn when marches to Hess's grave were banned. 22 By making it harder to mobilise and eventually quashing any chance of large demonstrations, authorities deprived far-right demonstrations of their audience, the attention of press and passersby. Maybe, too, the scrupulous separation of far-right demonstrators from belligerent counter-demonstrators removed what might have been an enticement to the far right: attempting to beat back leftist activists. Far-right adaptations failed to cope with these deprivations.
More broadly, the Hess Gedenkmarsch relates to other cases of far-right demonstration campaigns (not least later manifestations of the Hess Gedenkmarsch). Other cases exhibit the same causal conjunction of violent counter-mobilisation and state social control. While 'mechanisms have different empirical manifestations in different cases' (Beach & Pedersen, 2013, p. 73), we may presume broadly similar causal processes at play in other cases of this conjunction. The mechanism theorised in the first part of this paper manifests in other instances of campaign demobilisation. Further process tracing research would help to clarify the extent of similarity across cases.

Notes
1. NB: studies on repression reveal a complex causal nature: repression is, by turns, deterring and inciting. This suggests the effect of repression depends on the conjunction of repression with other causes. 2. See Staggenborg and Lecomte (2009) for a review of research on social movement campaigns. 3. A new Hess Gedenkmarsch campaign was mobilised in the early 2000s (and indeed again more recently); one may conceptualise it as one long campaign with periods of abeyance (e.g., Virchow, 2013b). However, in either conception, the demobilising mechanism remains: the chain of coercive counter-mobilisation followed by state social control resulted in a cessation of interventions, even if only for a number of years. 4. Earl's (2004Earl's ( , 2006 typology of social control circumscribes the universe of external sources of demobilisation. The typology relates to 'repression,' but Earl (2004, p. 58) favours the term 'social control' in order to avoid the misleading connotations of 'repression.' 5. In other words, are there other possible or even probable explanations given the observed evidence (Beach & Pedersen, 2013, p. 101)? 6. In other words, what must be observed to confirm the hypothesis (Beach & Pedersen, 2013, p. 101). 7. See Beach and Pedersen (2013) for a fuller explanation of the application of Bayesian logic to process tracing. 8. This contextual specification presupposes a state that protects freedoms of speech and assembly and places minimal restrictions on those freedoms. However, in other contexts the state may act with less neutrality. See Lasnier (2017), for example. 9. The definition of this concept is notoriously contentious; Mudde (1996) identified twentysix existing definitions. Without wading into this semantic tempest, it suffices to say that a widely-encompassing conceptualisation, such as Minkenberg's, is serviceable; the distinctions between variants of far-right social movements presumably do not give rise to a qualitative difference in the demobilisation of their large demonstration campaigns. 10. Minkenberg writes, there are 'four variants of radical right forces: (1) an autocratic-fascist right, usually involving racism or ethnocentrism and inspired by right-wing dictatorships of the interwar period; (2) a racist or ethno-centrist-but non-fascist-right, usually employing "ethnopluralist" arguments for the incompatibility of cultures and ethnicities while denying the existence of a "natural hierarchy"; (3) a populist-authoritarian right, organised around a strong and charismatic leader, with an authoritarian structure and a diffuse nationalist or xenophobic ideology; and (4) a religious-fundamentalist right, in which nationalism or xenophobia merge with religious rigidity, resulting in the defence of a religiously-framed conception of national "purity."' 11. That is, originally, 'Antifaschistische Aktion.' 12. On countermovements and counter-mobilisation, see Meyer and Staggenborg (1996), Meyer (2008), and McVeigh (2009). 13. The need for legal justification relates to the scope condition of 'localised demonstration campaigns' mentioned above. In cases where a demonstration campaign recurs in one place or region, justifying a threat to public safety can be based on direct past experience. This advantage is usually not available to state authorities trying to impose social control measures on non-localised demonstration campaigns. 14. Germany is one such context. Virchow (2007, p. 299) writes, far-right 'demonstration marches are no longer high-risk events. The NPD and neo-Nazi groupuscules can refer to decisions of the Federal Constitutional Court which has, on a number of occasions, lifted bans on extreme rightist demonstrations that had been imposed by town clerks' offices and confirmed by lower-level courts.' 15. In some instances, far-right organisers may employ wholly new tactics while still attempting to serve strategic ends, such as adopting smaller, community organising practices instead of large, planned demonstrations that are more susceptible to counter-mobilisation and state social control. While such tactical adaptation preserves the campaign, it marks a major shift. This can be conceptualised as non-demobilisation or continuation of the campaign, as opposed to negative demobilisation -the outcome of the mechanism theorised above -or 'positive demobilisation' (Davenport, 2015, p. 22), that is, demobilisation resulting from the accomplishment of campaign goals. 16. Other far-right groups participated later in the campaign, most notably the Freiheitliche Deutsche Arbeiterpartei (FAP) starting in 1989. The Hess Gedenkmarsch campaign was a coalitional demonstration campaign; indeed, a campaign that aimed to unify the German (and wider European) far-right movement, which was in some ways atomised into numerous uncoordinated groups (Virchow, 2013a). 17. As Vierkant (2015, p. 273) points out, there were several advantages to rallying around the figure of Hess, including previous experience that suggested Hess had a wide appeal among the far right and that far-right mobilisation might expect less resistance from anti-fascist groups than in Berlin. 18. For example, in the 'Dates' section of their publication in the first week of August 1990 (Nummer 1., 1990). 19. Tellingly, 'Info-telephones' first appeared in the Constitutional Protection Service's reports about right-wing extremism in 1993 (Bundesministerium des Innern, 1994). 20. That Antifa counter-mobilisation, too, adopted the tactic of telephone-coordinated convoys is in itself interesting, perhaps indicative of tactical immitation that is often observed between movements and countermovements. 21. On social movement abeyance, see Taylor (1989), Lapegna (2013), and Lasnier (2017) 22. This symbolic purpose that was given explicit attention in 2011, when state authorities exhumed Hess's remains and re-buried him at sea.