Social movements and political outcomes : why both ends fail to meet

The relationships between social movement challenges and political outcomes remain strongly under-researched in the field of social movements. Here, we use the labels “social” and “political” in a broad sense to comprise many types of challenges and many types of outcomes, such as economic and social outcomes for specific movements as well as general policy outcomes. Four theories are crucial for understanding successful mobilization of social movements: relative deprivation, resource mobilization, framing, and the theoretical figure of the opening political opportunity structure. Political outcomes, at least in democratic political systems, are usually the result of a parallelogram of different claims and means of influencing outcomes, in short, of compromises. Here, we list various forms of outcomes, from successful acceptance of movement demands to part-time successes or entire failures, and also the various strategies incumbents have in dealing with social movement challenges. Researchers usually have focused on the individual and structural conditions of the emergence of social movements but less so on the conditions of processing social movement demands and the outcomes for movements themselves, for the electorate and for policy changes. Consequently, there is little research 1 Portions of this paper were originally presented at the 12th World Congress of the International Sociological Association, Madrid/Spain, July 1990, and never published, though there were requests for it. On reading parts of the more recent literature, the author believes that many of the analytical arguments made are still to the point today. Consequently, he wants to share them with a broader audience, adding a few comments on recent social movements and their outcomes. It is up to the audience to improve on these arguments. Paper presented to IPSA RC 21-29 Conference on “Social and Political Movements, Leaders, Education and Communication. Challenges and Expectations in the Age of Globalization”, Lublin, Poland, September 11–13, 2014. EKKART ZIMMERMANN 32 available that would meet the requirements of an adequate research design in view of the numerous factors spelled out here as a theoretical control list. The idea of a response hierarchy of incumbents is suggested as a sort of a dispositional concept for further, more consolidated, research in this area. Also the notion of cycles of various sorts has to be kept in mind in order to avoid misjudging of both, the persistence of social movements over time, and their eventual successes and failures.


INTRODUCTION
That both ends fail to meet has been repeatedly noted by researchers prominent in the field of social movements and political protest. McAdam writes that "political scientists largely failed »to adequately explain or take account of the impact of social movements on the institutionalized political establishment«" [McAdam 1982: 2 as quoted by Tarrow 1988: 425]. Why both ends are not very likely to meet very often, is the theme of the present paper.
It is two generations since Gamson [1975] published his pioneering study on the effects of different social movement strategies and resources on the outcomes of social protest. Five years later, Gurr [1980] could present a broad summary of research on the outcomes of a violent political conflict. Yet, he also pointed to vast lacunae of research [Gurr 1980: 249-255, 291-292] very few of which have been closed in the meantime.
The confrontation between the theories of relative deprivation and resource mobilization theory has clearly been won by the resource mobilization theory. Relative deprivation requires awareness of one's situation and other groups to compare with. This notion adequately explains discontent resulting from social comparison processes. Within a polity the most deprived people, i.e., the absolutely deprived ones, often lack the means of making comparisons, because they are busy in their struggles for food and shelter in order to survive.
Yet, even massive relative deprivation does not guarantee mobilization of such discontent. Moreover, sometimes people are relatively more deprived outside a social movement that cares about mobilizing such feelings of deprivation. What is crucial is providing resources (e.g., personnel, money, time, skills, ideas and framing, information, networks) for setting up organizations and social movements and maintaining their drive. Further, not even resources mobilized are sufficient to explain the outbreak of challenges on the part of social movements. That is why the notion of the political opportunity structure is so important though it often comes in as a deus ex machina explanation. Rather it should be specified in advanced and accordingly be tested.
Also the plasticity of various resources is often inadequately assessed [Zimmermann 1999: 64]. Money, legitimacy, personnel or sometimes expert knowledge are mentioned as resources for mobilization. It remains unclear whether these are all the resources or only the more important ones. Rather the range of the various resources remains to be determined: their relative importance for the existence, the sequence and possible successes of specific types of organizations remains to be assessed in a differential diagnosis. Otherwise there is the danger of sampling on the dependent variable [Zimmermann 1999: 64].
Clearly mobilization is strongly affected by the presence of massive collective dissatisfaction and dissent. As such it comes close to a necessary factor for mobilization, but in itself is not sufficient for such mobilization to occur. A frame for interpretation and ideological commitment needs to be transmitted just as much as organizational structures have to be set up. Yet, if the intensity and extension of collective discontent remains un-assessed it will be difficult to assess the thrust of collective protest in a political opportunity structure that opens up. Developments in East Europe, in 1989, came as an equal surprise to East and West as to the speed of the mobilization and the spread of collective dissent [Kuran 1995].
The literature on social movements and political outcomes is widely scattered and rather uneven, with solid cross-national studies still comprising a small minority. On the one hand, this has to do with the difficulties of funding such studies and collecting data on the research questions involved. On the other hand, it is an indication of the underdevelopment of the field, abundant of single-case-analyses with little analytical perspective. Consequently, it comes as no surprise that leading scholars in the field come up with statements like the following ones. previous research, in particular survey research left unmeasured the "actual behavior as well as the interactions among protestors, opponents, third parties, and the state -in other words, the political process of collective action" [Tarrow 1988: 425]. "The interest of many scholars in social movements stems from their belief that movements represent an important force for social change. Yet, demonstrating the independent effect of collective action on social change is difficult" [McAdam, McCarthy and Zald 1988: 727].
Still, the 1980s and later decades brought enormous efforts to shift the attention to the political process, to the processing of political challenges. The works of Kitschelt [1986Kitschelt [ , 1988Kitschelt [ , 1989 on the rise of left-libertarian parties and Tarrow [1989] on protest and politics in Italy 1965-1975 stand out in particular. Kitschelt pursues a broad design and, at times, deals with as many as 18 countries when focusing on conditions (economic development, social security expenditure, strikes, left parties in government, intensity of nuclear controversy) influencing the presence of left-libertarian parties in the 1980s [Kitschelt 1989: 36]. Mostly, however, he focuses on carefully selected pairs of countries, e.g., Belgium and Germany [Kitschelt 1989] or the US, Sweden, France and Germany when dealing with the procedural and substantive impact of anti-nuclear movements on political regimes [Kitschelt 1986]. Tarrow, while focusing on a single country, nevertheless, starts from a clearly cross-national theoretical perspective and collects a vast amount of data. To the knowledge of this writer there is little such data for any other advanced industrialized country (but see Rucht 2012 and the PRODAT 2 data for Germany, 1950Germany, -2002.
These pioneering works once again demonstrate the strengths of a cross-national deductive theoretical approach, a perspective that is lacking for about four fifth of the literature published in the field of social movements and system response. The collections edited by Klandermans [1989] and Klandermans, Kriesi and Tarrow [1988], Tarrow and Della porta [2004] and in particular Della porta et al. [2009] are an indication that the field should move into the cross-national and theoretically deductive perspective, that there is a vast amount of case material available that needs to be cleared from a deductive point of view. The goals of the present paper point into the same direction. We want to 'clear some underbrush' in bringing together likely key variables accounting for the linkages between social movement challenges and political outcomes. Given what has been said above, any such effort must be considered preliminary. The focus is more on deriving fruitful dependent variables, on categorizing classes of potentially explanatory variables, on performing various analytical exercises on the base of informed ignorance and eventually setting up fruitful heuristic causal models that could guide further empirical research. Empirical references are given on an illustrative, not a systematic base.

TRACING SOME KEY VARIAblES
On the left side of any initial causal model one would find the challengers, a social movement, on the far right -the political (and social) outcomes, and in between -system responses.
social movement challenges ---> system responses ----> | OUTCOMES Since outcomes are influenced by numerous additional factors, many of which are even beyond the control of incumbents, not to say social movements, it seems to be a more fruitful research strategy at this point to concentrate on system responses as they relate to social movement challenges. Once the many facets of these prior causal interrelationships have become clearer, it may then be easier to widen the theoretical focus. Even from the rudimentary causal model in Fig. 1, however, it should be evident how far away social movement challengers are from the outcomes they desire. To simplify the discussion we have omitted a causal arrow going from system response to social movements. The fact that many social movements are a reaction to prior system decisions or "responses" will hardly be disputed.
The challenge of social movements is a function of • the goals pursued, whether anti-system (as in terrorism, since systematic use of violence is contrary to the state monopoly of violence) or within-system, • the resources available and used (organizational means, members and supporters, money, education, etc.) and Transatlantic Trade and Investment Partnership (TTIP), • the strategies and tactics selected from the repertoire of contention [Tilly 1978;Tilly and Tarrow 2006]. The system responses comprise various sets of actors, mainly • governments, • bureaucracies and the judiciary, • political parties, • the public, in particular the media, • and the population at large. Reactions within all five categories could occur at the federal, state and local level. "In a system of divided power, the legislative, executive, and judicial branches operate under different procedural and substantive norms and have different constituencies" [McAdam, McCarthy and Zald 1988: 721]. Rochon [1988: 194 ff.], e.g., provides wide evidence on judicial support for protest groups. In more centralized states, the set of actors is less complicated making perhaps for fewer political opportunities. In comparing implementation policies of nuclear power programs in Sweden, the US, France, and Germany, Kitschelt distinguishes between political input (open to access for social movements) and output structures (government capacity to implement). Where both are high, as in Sweden, anti-nuclear movements have been relatively successful. In France, e.g., input means are weak and output capacities high making for the effective and relatively undisturbed implementation of the French atomic power plants.
Since social movements (e.g., for the Equal Rights Amendment; pro-abortion movements; movements against the Transatlantic Trade and Investment partnership -TTIp; pro-immigration movements) often contribute to the creation of counter-movements (e.g., anti-abortion leagues; xenophobic movements like Pegida mostly in East Germany) the picture actually is much more complex. Moreover, different social movements compete for the same clientele of supporters, thus, letting a whole social movement sector come into existence [McCarthy and Zald 1977] with both cross-fertilization and increases of resources and deadly infighting taking place.
perhaps the most fruitful theoretical development in the field of social movements and political outcomes over the years has been the notion of political opportunity structure. It is present in such diverse theoretical works as diverse as those of Eisinger [1973], Gamson [1975], piven and Cloward [1977], Tilly [1978], Gamson [1975], McAdam [1982], Tarrow [1983], Jenkins [1985], and Kitschelt [1985Kitschelt [ , 1988Kitschelt [ , 1989. Two notions seem to be prominent here: first, social movements need to form coalitions with "members of the polity" [Tilly 1975: 547] or to win support from such groups, in particular if they aim at major goals. Second, and more important, it is less through the efforts of social movements, but rather through openings in the political structure that the likelihood of goal achievement is increased. In this respect piven and Cloward [1977] have stressed the factor of electoral instability. Other authors have emphasized additional factors such as "broad shifts in public opinion" [Jenkins 1983: 547], elite disunity [Tilly 1975] or widespread disturbances [piven and Cloward 1977;Tarrow 1989]. "When favorable changes in public policy and legislation occur, the social movement input is one component of a larger coalition of political groups, lobbies and supportive publics, all of which were crucial for success" [Oberschall 2012: 189].
Depending on whether openings in the political opportunity structure are favorable to social movements, the potential set of reactions of incumbents and other pro-system forces is either left unchanged, diminished or even increased. Later on, we shall return to this notion of the response set (not to be mixed of up with its meaning in survey research). Here, for the moment, the set of reactions available to governments and other actors on the right side of Fig. 1 should be briefly listed. Social movement challengers often forget about this broad set of responses open to incumbents, especially to broadly legitimated governments.
Analytically, government responses (and strategies) can be grouped into two categories: a) denial and b) acceptance of the problem. There is a third category, namely to ignore a problem, that logically cuts across this basic distinction. Ignoring a problem could mean both, denying its existence or conceiving indifference as one of the responses to the problem, maybe till further information makes action more urgent. Chancellor Merkel`s (mis)handling of the recent immigration wave and following internal and international repercussions is a striking example. In democracies ignoring a problem on the part of incumbents and the media seems to be a much more effective strategy than flatly denying it. Chancellor Kohl was a master strategist in this realm. This can be widely illustrated, e.g., how the nuclear rocket deployment issue was handled after 1982 or how he chose to ignore interview invitations of the leading political weekly, "Der Spiegel", instinctively sensing that he might be a loser whatever he is going to say. Challengers often underestimate the fact that ignorance can be a powerful weapon for broadly-legitimized incumbents.
To return to the basic alternative: denial of a problem. Again, two elementary responses exist: using counter-information that would turn a "problem" into a "non-problem" or even a "blessing", or relying on repressive strategies such as criminalization of social movement actors or the employment of brute force.
In responding to the problem, no such clear-cut dichotomies seem to emerge. Rather a wide array of tactical measures and strategies seems to be available to win back or retain an electorate that could be tempted by new social movements. Among such strategies of issue competition are: • "buying" the issue and transforming it into a major issue high on the political agenda; • embedding it amongst other issues, thus, placing it lower on a response hierarchy; • "neutralizing" it through the dominance of a new issue (e.g., ecological questions in West Germany through matters of German unification). Even these rudimentary classifications make it clear that the resources of social movement challengers usually are limited vis-à-vis incumbents and other pro-system forces. "Advantages [for social movements] are straightforward for the short runspecific gains on voting rights increase voter participation and translate into elected officials -but long-term impacts of collective action are difficult to measure because conflict is a dynamic system and the adversaries will react to temporary setbacks" [Oberschall 2012: 189].
Ignoring the special theoretical claims of piven and Cloward [1977] of unruly behavior as the only means of poor people to achieve (part of) their aims (the costs of organization are said to be too high for these groups) and the criticism they have met [see Jenkins 1983: 545 for references], the threat potential of social movements to incumbents could be scaled as follows: 1. Social movements have to compete with or work in close cooperation with political parties. As pointed out by Smith [1976], this is the only chance how challengers via institutional means can enforce a zero-sum game on incumbents. political protest movements rarely inflict lasting damage on other more established parties, since protest groups infrequently participate in elections. In contrast to other forms of political competition, electoral votes imply a zero-sum contest. Beyond the polls, the more established groups and the state command multiple resources to out-manoeuvre their challengers. Taking away voting resources from incumbents takes away also other resources from them. Voting strength in turn is a function of holes in the issue space as well as the party system space [Kitschelt 1989]. If social movements can operate from a monopolist position in issue areas of growing importance, their electoral chances are rather good. 2. If such protest parties take electoral hurdles and enter parliament, they usually have to find allies within parliament. Most likely they will find those, if at all, amongst other opposition parties. Depending on the coalition calculus, however, former protest parties could turn into becoming pivotal parties (as e.g., the Green party in West Germany after the state elections in Hesse in 1983 or in spring 1990 in lower Saxony). In general, however, one would expect them to spend some years in opposition before they will enter step three. Tilly [1975] has developed a broad theoretical scheme that makes crucial distinctions whether challenging groups find allies within the system forces, in particular on the part of dissenting elites. 3. The biggest hurdle on the way to achieve the policy goals of social movement actors is joining a government coalition. Yet, even if in such a position party competition continues, and challengers might be utilized only as helpful "idiots" to secure parliamentary majorities but to be dismissed at the next political opportunity (e.g. the Communist Party in France by President Mitterrand). 4. Moreover, turning to policy outputs, they frequently are co-determined by supra-national institutions such as the EU or requirements of the world economy, the environment in general or by special events. There is also an international nesting of political opportunities going beyond local opportunity structures and national opportunity structures [Meyer and Reyes 2013: 222]. The anti-IMF food riots tied to the calls for austerity programs on the part the International Monetary Fund (and the World bank) bundled "widespread public opinion against the impending economic reform, large-scale mobilization by multiple social sectors, and a strong oppositional political party acting as a friend inside and outside of the polity" [Almeida 2012: 317]. Globally-induced protests such as those against TTIP make for further variants of political opportunity structures, beyond the local and national scenarios [cf. Tarrow and Della porta 2004;Tarrow 2005;Della porta et al. 2009]. Yet, even when reaching step three, nothing guarantees that those policy outputs that were originally envisioned by movement activists will be achieved. This discount list of hopes and myths is ended with a final distinction: 5. policy outputs are never to be mixed up with political outcomes. We speak here of political outcomes in a broad sense as to comprise measures of social change, since most of those changes are mediated by other political decisions as well. Of course, social movements could opt for unruly behaviour, but long-term achievements in advanced liberal democracies are most unlikely to be achieved through such means. The failure of Italian and West German terrorism and other non-ethnically motivated terrorism in advanced industrialized societies clearly attributes to this.
Does the vulnerability of social movements decrease as they march on the scale from 1 to 5? More likely than a linear relationship that would specify a decrease in vulnerability as they enter parliament with broad political support, is a curvilinear relationship. If Smith [1987] is right and if a zero-sum game is indeed taking place, by definition the vulnerability of social movements (or new parties derived thereof) increases as they enter the race of parliamentary competition. Part of this argument is brought forward, though often with different intentions, on the part of fundamentalist social movement leaders (e.g., green fundamentalists in several countries) who want to keep a movement going rather than opt for organizational factors and truly parliamentary competition.
Our scale is similar to others in the literature, e.g., that of Huberts [1989: 407] who conceives of the following hurdles: wants ---> demands ---> issues ---> decisions ----> outputs Frequently, however, the sanction potential of movement challengers is drastically reduced through such factors as • sectarian infighting, • "rioting for fun and profit" (e.g., during the Arab Spring) and other likewise activities entertaining or scaring the public, but destroying resources and appalling potential sympathizers, • the structure of single-issue movements running the risk of being bereft of their focus of crystallization [Zald and Ash 1966], or • electoral clauses not met. These factors could operate on an additive (or substitutive) as well as on a multiplicative base [for further arguments here cf. the works of Tilly 2004, andTarrow 2006].
With their command of large portions of the state apparatus (e.g., judiciary, police forces) the incumbents thus have a tremendous advantage in resources over social movement challengers. The "criminalization" of terrorist challengers hands them over to the judiciary, before any political debate is opened. At the latest point, terrorists in West Germany, Italy and elsewhere learned how fateful it was to underestimate the strength of the resources available to the authorities and how efficient a division of power (and labor) can be once those resources (e.g., repressive resources) are used, in particular if backed by large portions of the population.
The short-term challenge might be enormous, if the media and public attention focus on terrorist attacks or social movement claims. In mid-term and long-term perspective, however, this is far from the zero-sum game institutionalized through competition at the voting booths. In using examples from terrorism and the social movement sector at the same time, we do not want to juxtapose the two political phenomena. There are important differences: social movements have a larger following, they address concrete issues in a mostly legitimate way. The opposite is true of non-ethnic terrorism in the mentioned countries. Analytically, however, some of the same conclusions might have to be drawn, depending on what is under consideration.
Given the general imbalance in resources and the long march to political power on the part of social movements, it comes as no surprise that some of their biggest achievements lie in the refinement of the repertoire of action. In inventing and unfolding new forms of collective action and civil disobedience and in broadening the public discourse (agenda setting), social movements fulfill two of their major functions often in a masterly way that contrasts so much with the modest achievements when it comes to political outcomes envisioned. Unconventional political protest and large portions of social movement challenges in their role and function are succinctly captured by drawing on the analogy of binoculars. Social movements can draw public attention to problems very efficiently, in particular if inventing new tactics and broadening the repertoire of contention. Yet, if the political wanted outcomes are in focus, the binoculars are turned around causing a wide-distance-effect. And this is often true in a double sense: first, the solutions suggested on the part of social movements are dictated by single-issue considerations leaving out opportunity costs. Thus, they frequently are unable to gain the necessary support for a political compromise. Second, and more important, both ends of the binoculars cannot meet, on the base of the previous discussion and further arguments below.
Those analytical assessments to be specified below should not distract from the many functions social movements have in liberal democracies. Among these, some are drawing attention to neglected issues and bringing about social change, protecting minorities against the tyranny of the majority and providing a training ground for new political leadership, all of which are more likely to increase the functional integration of the society and polity.

IN SEARCH OF A DEpENDENT VARIAblE
Given that the binoculars are turned around the more we move over to the right side in Fig. 1, what are then dependent variables that make for fruitful comparative analyses as to the achievements of social movements? The bigger the goal (and most likely thus the bigger the challenge), the less likely total goal achievement. Gamson [1975] used a crude dichotomy: group acceptance from its antagonists and new advantages (e.g., favorable law-making, change in public beliefs), neglecting here the groups that never made it into system acceptance. Yet, his sample stops in 1945, in a world rather different from today's world-wide linkages, big bureaucracies and administrations that (co-)determine everyday life conditions in advanced liberal democracies. Also, several of his classifications and his statistical analysis have been debated in the literature [cf. e.g. Goldstone 1980; Gurr 1980].
Outcomes could be distinguished for • ordinary movement members, • movement elites (e.g., new administrative positions vs. gliding down on the social stratum, if a movement fails), • the population at large (issue attention), • counter-movements and other groups that feel threatened through the movement challenge and • the incumbents. Having in mind some major social movements of the last one hundred years (leaving out fascist movements), at least the following criteria of long-term success could be mentioned (in declining order): • structure of the goal (reconcilable with the prevailing system of productive forces, non-zero sum conflict structure), • economic bargaining power/organization power, • voting resources/and organizational efficacy, • (liberal) allies. In terms of these four criteria the Labor Movement is the most successful social movement. Yet, even here in inter-war Europe or in post-WW II Eastern Europe, periods of severe backlash occurred. The Civil Rights Movement would score positively on the structure of the goal and liberal allies, but could not command over economic bargaining power or strong electoral resources [cf., e.g. McAdam 1983 on the decline of this movement].
The Student Movement scores only on liberal allies, but its fuzzy structure of goals and lack of economic bargaining power and voting resources contributed to its eventual demise. A similar analysis could be performed for the movement for nuclear disarmament of the 1950s and 1960, only that the structure of the goal was clear, but much more difficult to communicate and less likely to lead to a realistic challenge of existing super-power threats and consequences derived thereof. Yet, the long-term disarmament process between the West and the Communist bloc could, to some extent, be attributed to the successful challenges of the Peace Movement (Helsinki process).
Political scientists have focused on the formation of new political parties, e.g. left-libertarian parties [Kitschelt 1989;Müller-Rommel 1985] and their political performance. While such analyses underscore many of the theoretical points made above, they, nevertheless, clearly move away from the focus on social movements, their pursuance of goals, the interactions in the political opportunity structure and the analyses of eventual outcomes. As a starting point, however, it seems to be a wise decision to go for such a dependent variable since it interlocks with both the institutionalized realm of politics and the challenges brought up by new social movements [or sometimes rejuvenated social movements; cf . Taylor 1989;Fuentes and Frank 1989: 179-81].
Yet, explaining the rise of these groups and thus, at least, enhancing our knowledge as to the channels and linkages to and from social movements to party politics is a much more difficult task. Thus, Kitschelt [1989: 17-40] finds "reasonable" empirical evidence for economic development to have an impact on "successful left-libertarian party formation". At the same time, he dismisses two other explanatory variables, the "relative size of the student population in advanced education" and the "size of the tertiary occupational sector".
As clear such a dependent variable is, as difficult is the linkage backwards to social movements and forwards to outcome assessments. Moreover, a solution of problems addressed by social movements is admittedly rare. This would do away with the need of these movements. Thus, one has to ask why social movements do not persist over time. Since most problems are not solved in a way social movements could be happy with, the answer must either be sought in the formation of new parties which take over or party wings or interest groups, or in the cyclical composition of much of social protest and social movements [Fuentes and Frank 1989]. linear models and equilibrium models often inherently used in social science may be inadequate for the task ahead.
Yet, before turning to the cycle argument again, there is perhaps another theoretical solution to the problem of an "adequate" dependent variable in research on social movement outcomes, although our discussion will be set at a rather abstract level. If, however, the argument should prove a valid one -in the sense of a "dispositional concept" -the research strategy on the outcomes of social movements might indeed have to be severely changed.

ON RESpONSE SETS AND RESpONSE HIERARCHIES
Assume that governments, administrations and other actors entrenched in the system have a response of options available, a response set. Some of these options are limited and are exerted in a strict stimulus-response pattern (e.g., appointing substitutes upon the killing of political authorities). More generally, however, it is not implausible to conceive of a set of broad answers available in a hierarchy of responses. (The same incidentally is true of social movements, and has been widely studied under the terms of refinement of the strategic arsenal, but not under the present theoretical perspective.) One measure to score the "success" of social movements with respect to the state's and other actors' responses is whether they change the response set pattern, whether a reaction lower in the hierarchy is moved up by the challenge of a social movement. Budgetary changes, e.g., are made in response to urgency needs. On the other hand, the failure of West German terrorism could easily be illustrated in that the big change wanted, namely an over-repressive reaction on the part of the authorities, did not take place. Though repression was somewhat increased, this response in the response set of measures was not moved up so much to the front as hoped for in the terrorists' strategy. Rather the temporarily increased level of political repression was suspiciously watched and monitored by the liberal public.
There are several caveats and problems of observation raised through such a conception. First, some responses in the hierarchy may be independent of the challenge, e.g., those responses determined by constitutional provisions of divided power. Second, there may be institutional fiats and rules not allowing for any change of a response in the hierarchy. Third, some responses would have occurred irrespective of a challenge (e.g., welfare measures to be passed anyway). Fourth, governments may choose not to react at all or only to react to the means (e.g., violence) selected, and not to the goals proclaimed.
The responses may be dictated by other problems or problem arenas, by a retaliation threat, by loss of public support or fear of losing an election, and simply by personal courage moving up a reaction on the ladder. If such a hierarchical list of responses exists, it would be influenced by, e.g., (a) the structure of law-making and institutional differentiation, (b) by the traditional use of certain measures and responses, (c) by public support for specific measures, and (d) by (additional) political calculations in given circumstances.
Yet, how to determine such a response set? Specific challenges, goals, problem areas in a number of countries would have to be analyzed, both over time and in cross-national perspective, to tell us something about the setup of those response hierarchies, their change and actual use. What we have in mind here may be available only at the end of a gigantic research endeavor. Yet, it also could prove a helpful starting point for carefully selected paired historical comparisons (or groups of pairs) where at least some variation in other independent variables is controlled for (see Ragin 1987 on many examples and considerations for such research designs).
A notable change in the response hierarchy, moving a measure down or up (e.g., welfare regulations on the part of the labor movement), could then be treated as a measure of the dependent variable (still leaving the problem of outcome analyses). We are aware that a simpler design going for the extension of welfare measures and relating it to the strength of labor votes, has parsimony on its side, and lots of results to report. Yet, in terms of a more coherent, broader and yet specific theoretical explanation, the present notion of changes in the response set might prove useful. Sceptics might say that through such a strategy both ends of relating social movement goals, resources, and strategies to political outcomes will never meet. At least the debate should be opened. It might, however, be quickly closed again when considering the strong empirical evidence brought forward in a body of theorizing that deals with cyclical arguments.

CYClES OF pROTEST: HOW DEADlY AN ARGUMENT?
The cyclical argument comes in many variants. Sometimes this leads to overlapping streams of argumentation, sometimes these arguments supplement each other. First, there is the birth cohort argument [Easterlin 1980]. Dense cohorts experience more strain on the labor market and during lifetime, whereas there is a great demand for those cohorts that are smaller in numbers. Small numbers should have a positive impact on personal welfare. Even though many individuals of larger cohorts on individual rational choice abstain from protest and try to double their personal efforts, it is not unlikely to associate more political and social turmoil when larger cohorts are raising their voice in society. This holds for the student revolt. On the other hand, there are few cross-national and diachronic studies that systematically tackle the relationship between strength of cohorts, social movement growth, and political and social protest behaviour [see Urdal 2012 for one].
Second, Hirschman's [1982] shifting involvements -from public political engagements and the frustrations experienced therein back to private lives -are wellknown. Whereas in the first case, changes between different age groups are to be examined, here changes over the life-cycles within the same cohort are at issue. Thus, there may be a cycle involved within the broader generational cycle outlined above.
Third, and related to the previous point, there are the costs of mobilization and demobilization [Tilly and Tarrow 2006]. These costs occur irrespective of goal achievement. With goals failed (or partly achieved), however, they become even more burdensome and likely to make for periods of abstention after social movement peaks. One such peak with a subsequent decline occurred in the first half of the 1980s in Western Europe, in particular in West Germany, when the peace and anti-nuclear movements coincided with the ecological movement and portions of the women's movement and other movements as well.
Fourth, and most important thus far, cyclical political shifts are built into Western democracy. The cyclical occurrence of elections gives the former opposition a chance to take over government. The political leeway, the openings thus created in the party system seem to be a major force in propelling social movements and new political parties. Alber [1985] has argued that labor parties that bid farewell to the ideologies of class conflict and happen to be in (or part of) government, create political opportunities for groups of leftwing sympathizers, in particular unemployed or under-employed academics. On the far right of the political spectrum, there are related examples. In West Germany, e.g., those social movements and political forces seemed to be particularly challenging when the Christian Democratic and Christian Social Parties were in government thus not being able to bundle political and social dissatisfaction of the political extreme right as it would have been possible in the role of the major opposition force. The political opportunities created through the link: elections ----> party system change, have been most intensely studied. This body of research represents the best corroborated findings in comparative research on political opportunity structures for new social movements. "More broadly, political opportunities affect the social movement's potential to mobilize, advance particular claims, cultivate alliances, employ particular strategies and tactics, and influence mainstream institutional politics and policy (…). The puzzles remain, however, as openings that encourage voters mobilization sometimes line up with opportunities for policy influence, yet other times, protest opportunities occur precisely when the prospects for policy influence is most distant (…). This is, at least partly, a function of whether activists mobilized in response to prospective gains or prospective losses. In the latter case, 'winning' may comprise little more than stalling unwanted changes" [Meyer and Reyes 2012: 221].
Fifth, such electoral cycles do not seem to be unrelated to both, protest cycles and reform cycles, the latter two, in particular, being the theme of Tarrow's [1989] work on Italy. "There were three features of this diffusion process that bear underscoring: first, disruption began institutionally in the context of conventional organized protests and strikes; second, the protests gained notice when 'early risers' broke through the constraints of convention during a brief intensive peak of mobilization; third, the peak triggered a long gentler cycle when others, less courageous but more numerous, saw that the system was vulnerable to protest and used institutional channels to forward their demands" [Tarrow 1989: 338-9]. Violence never predominated; it became, however, more general towards the end of the cycle. The pattern of the protest cycle allowed elites to "eventually segment the movement by a strategy of piecemeal reform and repression" [Tarrow 1989: 339], but also led to the acceptance of organized labor, to marked changes in voting patterns and other "voice" options.
What follows from these theoretical observations for the study of social movement outcomes? If those challenges occur in cycles and are tied to reform cycles as Tarrow argues, then we have to opt for cross-national research designs that will bring in explicitly those cycles with both peaks and latency (non-event) phases. Yet, such a longitudinal design also increases time and space coordinates making it more difficult to control the variation of other factors. Where are the cutoff points of those cycles? It would be a Herculean task to provide similar data sets for other countries as Tarrow has amassed for Italy. (It took his team nearly a decade to do so.).
The other alternative would be to study only those social movements and their outcomes which do not seem to be affected by the cycles of elections, protest and reform (and maybe the birth cohort cycle). Yet, again and coming full circle, how do we know that appropriate cutoff points for studying those particular movements and their outcomes have been selected, when the full cycle is left out of consideration? Is a study like Kitschelt's [1989] which focuses on a clearly delineated dependent variable, the rise of left-libertarian parties, the maximum we could expect in terms of a coherent cross-national design? Or are there ways and means to follow the in-depth-strategy of Tarrow in concentrating on "all" protest events in a period that most likely covers such a cycle? CONCLUSIONS Broadening space and time coordinates in cross-national research on social movements and social movement outcomes is a gigantic task that eventually has to be tackled. Oberschall mentions conflict dynamics as new directions for research. Issue accumulation results from the sequences of conflict interaction and "explains the protracted character of many conflicts" [Oberschall 2012: 190]. Also collective myths and group solidarity may enhance the mobilization of discontent but also hinder more permanent settlements.
It seems as if there were two issues involved, one being how to collect all those data, the other being how to analyze those data given the great variation in time and space coordinates? Ragin [1987] provides many hints how one could arrive at meaningful conclusions even with "fuzzy" data [see also Shadish et al. 2002]. As much as Kitschelt's study is to be commended, is it really the maximum one could expect with Pandora's box of wide time and space coordinates open? Movement rises and strategies can be explained as well as some of the effects they have. Yet, a distinct assessment as to their contribution to broader outcomes, at present, seems to be beyond the capacity of individual researchers. Apart from what makes people voice their protest, this is the single most important question to address: what is the place of political protest, what does it achieve and how? "The 'new wars' […perhaps not the most adequate label, E.Z.] and social movement (…) theories come at conflict from opposing directions. [New war] theory starts with states and regimes as its unit of analysis, and descends to ethnic groups, political organizations, and insurgents. [Social movement] theory begins with small groups, networks, crowds, leaders, activists, and ascends to ethnic groups, social movement organizations, and regimes" [Oberschall 2012: 189-90]. This observation may be appropriate when primarily the level of analysis and thus the unit of analysis is at issue, i.e. when there is more emphasis on macro-dynamics or on micro-dynamics. States are macro phenomena, social movements start from individuals, groups and networks and then, as challengers, may turn into macro challengers of a state or of vital institutions.
The Occupy Wall Street Movement is one such latter example. It also exemplifies how difficult the attribution of outcomes to a social movement is. The movement has almost entirely vanished, probably only marginally affected the states' reaction to mismanagement in the banking and financial sectors but perhaps more positively acted as some form of a catalyst for new and harsh control measures as well as "financial repression". Too many other issues at the global security level or various local scenarios were important and salient at the same time. Thus, it would be premature to attribute the success in establishing new financial regulations and enacting draconic penalties against individual banks and financial leaders alone to the protests issued by this Anti-Wall-Street movement whose predominant social substratum mainly came from above modal values in income and education. A clear series of demands successfully met are those in the consumer realm when consumers react to bad food or exploitative conditions of production. Yet, frequently and after a short while the next scandal occurs which is another indication of "market flexibility".
If one turns to political and social movements challenging the political system as such, in short revolutionary challengers, the failure of the protests for democratization in the Arab world comes to mind as another series of events that links highly complex types of political conflict. Zimmermann [2013;2015] tries to provide both: first, a differential analysis of the Arab Spring vis-à-vis comparative revolutionary analyses with the four dominant outcomes (revolutionary success, repression, civil war, or succession), and second, to account for the causes and circumstances of failure in the Arab context. Here sultanism as a form of political system plays a large role. Following linz and Stepan [1996], sultanism is unable to provide established rules and channels for a (peaceful) transition to another political regime type. The lack of the state monopoly of violence in many Arab countries, coupled with inefficient economies, and widespread corruption, together with the youth bulge and the lack of a regional power keeping order are other major factors here [Zimmermann 2015].
If the binocular analogy proves correct, and we see no major challenge at this point, one could wonder about the practical implications of this type of social movement research. It is one thing to sympathize with certain social movements. It is another to have your own results being influenced by such a predisposition as often has been the case in research on social movements. It is a third, and more general point, to keep democracies efficient, to strengthen the input into it and to try to reduce the waste of political and psychological energies that takes place when there are misconceptions about social movements and their impact. Increasing our knowledge in this domain would save lots of disappointment on all sides respecting the rules of democracy.
practical issues are thus involved, as much as the needs of scientific curiosity and the pursuit of knowledge are to be satisfied in a professional manner. pandora's box is open; most wanted now are researchers grinding their teeth! bIblIOGRApHY