Skip to main content Accessibility help
Hostname: page-component-55597f9d44-mm7gn Total loading time: 0.803 Render date: 2022-08-16T22:47:33.716Z Has data issue: true Feature Flags: { "shouldUseShareProductTool": true, "shouldUseHypothesis": true, "isUnsiloEnabled": true, "useRatesEcommerce": false, "useNewApi": true } hasContentIssue true

1 - Introduction

A New Approach for Studying Political Contention – Contentious Episode Analysis

from Part I - A New Approach to the Analysis of Contentious Episodes

Published online by Cambridge University Press:  05 November 2021

Abel Bojar
European University Institute, Florence
Theresa Gessler
University of Zurich
Swen Hutter
Freie Universität Berlin
Hanspeter Kriesi
European University Institute, Florence


On 24 May 2011, in the middle of the parliamentary debate on the so-called mid-term adjustment plan, yet another round of austerity imposed by Greece’s international creditors, a call for a demonstration at Syntagma Square in Athens and at the White Tower in Thessaloniki appeared on Facebook. By the next day at least 20,000 people assembled in the two squares, mostly chanting “thieves, thieves” at parliamentarians and cursing the Parliament. The movement of the Greek Indignados or Aganaktismenoi was born. It would prove to be massive, expansive, and innovative. Immediately after the initial demonstrations, the main squares in the two cities were occupied, and simultaneous protests began in almost all major urban centers of the country. Interest would focus on Syntagma Square, however, where the occupation was symbolically confronting Parliament, juxtaposing the public assembly and the symbolic seat of political power. In the following days, the occupation grew exponentially, eventually reaching almost 400,000 participants on June 5th. In our dataset, there is an event associated with the Aganaktismenoi on almost every single day until the end of the episode on June 30th.

Contentious Episodes in the Age of Austerity
Studying the Dynamics of Government–Challenger Interactions
, pp. 3 - 23
Publisher: Cambridge University Press
Print publication year: 2021

On 24 May 2011, in the middle of the parliamentary debate on the so-called mid-term adjustment plan, yet another round of austerity imposed by Greece’s international creditors, a call for a demonstration at Syntagma Square in Athens and at the White Tower in Thessaloniki appeared on Facebook. By the next day at least 20,000 people assembled in the two squares, mostly chanting “thieves, thieves” at parliamentarians and cursing the Parliament. The movement of the Greek Indignados or Aganaktismenoi was born. It would prove to be massive, expansive, and innovative. Immediately after the initial demonstrations, the main squares in the two cities were occupied, and simultaneous protests began in almost all major urban centers of the country. Interest would focus on Syntagma Square, however, where the occupation was symbolically confronting Parliament, juxtaposing the public assembly and the symbolic seat of political power. In the following days, the occupation grew exponentially, eventually reaching almost 400,000 participants on June 5th. In our dataset, there is an event associated with the Aganaktismenoi on almost every single day until the end of the episode on June 30th.

At first, the Pasok government reacted to the movement in a mix of fear and embarrassment, but the original ambivalence soon gave way to growing anxiety. On the eighth day of the occupation, Prime Minister Papandreou addressed the ongoing mobilization, attempting to shift the blame to abstract “global powers” – to no avail. On the same day, the movement blocked all the exits from Parliament, effectively locking the M.P.s inside. Eventually, the M.P.s had to escape in the dark, with the help of the fire brigade, through the adjacent National Garden. Pasok M.P.s were becoming the main target, bearing the brunt of the opposition to a policy about which they themselves had considerable reservations. They reacted by challenging the government, asking for explanations and for assurances that this austerity package would be the last one. On June 7th, in the parliamentary committee, the five ministers in charge of the bailout took fully thirteen hours to convince raging and fearful M.P.s of the need for new measures. The protestes in the squares, which were initially seen as a potential relief for Pasok, were by now fissuring the link between the government and its MPs.

At the same time, EU pressure on the government escalated, as did its pressure on the opposition leader to share responsibility for the new measures. The opposition, however, did not budge. On June 14th, when one Pasok M.P. resigned and another publicly declared that he would not vote for the midterm adjustment, the government majority shrank to only four M.P.s. At this point, the possibility of a lost vote and a subsequent chaotic default loomed large. At the same time, the unions entered the fray. On June 15th, the large strike demonstrations of the unions fused with the Syntagma Square occupation, gathering hundreds of thousands once more. The earlier blockade was repeated. For the first time after twenty days of protest, the riot police moved in forcefully to disband the blockade, and the new movement underwent its baptism of fire. Reports of police repression and brutality carried out on a crowd that was until then peaceful shocked the attending public.

On the evening of the same day, the prime minister called the opposition leader to ask for a government of national unity. The latter accepted, on the condition that the government’s sole focus would be the renegotiation of the bailout and that elections would then be called. Papandreou first agreed but then withdrew his consent to such a program within a day and, instead, opted for a cabinet reshuffle, replacing his finance minister who had been the main target of the ire of the protesters and M.P.s. The new minister tried to open a dialogue with the movement and the unions. After the reshuffle, the government asked for a vote of confidence, which it received on June 21st. After having been finalized in the various committees, the midterm adjustment package was introduced in the plenary session.

The major unions responded with a forty-eight-hour general strike on June 28–29, the days of the plenary debate and the final vote on the program. The demonstration on June 29th, attended by both the unions and the Aganaktismenoi, proved to be one of the largest to date. While each organization had its own bloc, radical left parties, anarchists, a loose nationalist crowd, and Indignados united their forces in the first showing of an informal anti-bailout coalition. During the following night, while the midterm adjustment was legislated, a large group of hooded protesters clashed with riot police. As it turned out, the cabinet reshuffle and the signal of the new finance minister that he would consider social concerns sufficed to relieve the tension within Parliament and allowed the remaining Pasok M.P.s to vote compactly in order to pass the midterm adjustment program on June 30th. External pressure had vanquished the domestic threat from the new challengers. After the passage of the bill, the challenge subsided. The combination of repression and unresponsiveness by the elites deflated the movement.

This sequence of events, which has been told by Altiparmakis (Reference Altiparmakis2019: 143–154) in more detail, dramatically illustrates the patterns of interactions between challengers (Aganaktismenoi, unions, anti-bailout coalition), the government, and third parties (M.P.s of the governing party, opposition parties, foreign creditors) that had been triggered by austerity proposals by European governments during the Great Recession, one of the great economic crises in our time. In this book we shall study such patterns of interaction in twelve European countries. The Great Recession, which was unleashed by the breakdown of Lehmann Brothers in fall 2008 soon spilled over to Europe, where the initial shock of the financial crisis was to be followed by the Eurozone crisis, initiated in early 2010 with the sovereign debt crisis in Greece. While the worst of the crisis seemed to be over by fall 2012 after the head of the European Central Bank had declared that he would do “whatever it takes”to save the euro, the fallout from the crisis continued to haunt Europe at least until the conclusion of the third Greek bailout in summer 2015. It is hard to overstate the sheer magnitude of the impact the economic crisis has had on the lives of people in some, although not in all, parts of Europe. As Adam Tooze (Reference Tooze2018: 5) has observed in the introduction to his account of “the first crisis of a global age,” the combination of these crises and the economic and political responses to them are essential to understand the changing face of the world we are living in today.

Initially, governments countered the economic impact of the crisis by relying on some version of “liberal” (Pontusson and Raess Reference Pontusson and Raess2012) or “emergency Keynesianism” (Hall Reference Hall, Kahler and Lake2013). Once the Greek crisis deepened, however, starting in early 2010, governments turned to austerity policies, which were the key sources of economic hardship in the most hard-hit countries. While the welfare states buffered the negative consequences of the crisis initially (Bermeo and Bartels Reference Bermeo, Bartels, Bermeo and Bartels2014), especially in the countries of northwestern Europe, which had strong automatic stabilizers, the turn to austerity impeded the redistributive functions of the state and crucially contributed to the hardship of the populations. This aspect has focused the minds of the challengers on government policy and on the supranational constraints imposed on the national governments by agencies of the European Union, fellow governments of the Eurozone, and the I.M.F.

In the present volume, we focus on the interactions between the governments and their challengers in reaction to the governments’ austerity proposals. We examine the austerity proposals and ask whether and how they have been challenged by social movements, unions, opposition parties, and other actors, and how the governments, in turn, have reacted to such challenges. We are trying to understand how it was possible that austerity came to pass in spite of popular resistance by investigating in detail the contentious episodes that were triggered by the austerity packages proposed by European governments. As we shall see in the subsequent analyses, the Greek episode that we used to illustrate the interplay between challenger actions and government reactions is an extreme case with regard to the contentiousness of the challenge and the intensity of the interaction between the two main protagonists. However, it proves to be rather typical with respect to its outcome. Even in a case of a very intense challenge such as this one, where the government was heavily shaken by the mobilization of the challengers, the authorities ended up imposing their austerity proposals. We shall try to make sense of the patterns of interaction by analyzing in detail the composition of the main protagonists, how they reacted to each other, and the extent to which their reciprocal reactions depended on contextual conditions.

In focusing on the patterns of interaction that developed during the contentious episodes unleashed by the austerity proposals of European governments during the Great Recession, we believe that we can achieve a better understanding of what happened during this crucial period of European politics. We already know that the crisis has been particularly deep in southern Europe, where it led to a wave of public economic protest against the government austerity programs, while protest remained much more limited in northwestern and centraleastern Europe (Kriesi et al. Reference Kriesi, Lorenzini, Wüest and Häusermann2020). We also know that the southern European party systems have been profoundly transformed by the electoral consequences of the Great Recession, while the party systems in the other two European regions have been more resistant to change (Hutter and Kriesi Reference Hutter and Kriesi2019). However, our knowledge is based on a comparative-static analysis, and we have little understanding about the processes that have shaped the waves of protest and the electoral outcomes. It is the ambition of the present study to dig deeper into the dynamics of these processes in order to show the mechanisms that have driven the different outcomes at the macro level in the three European regions.

We have selected twelve countries to study the patterns of interaction underlying the macro-level outcomes – four countries each from the three regions of Europe: France, Germany, Ireland, and the United Kingdom in northwestern Europe; Greece, Italy, Portugal, and Spain in southern Europe; and Hungary, Latvia, Poland, and Romania in centraleastern Europe. In this selection we find several countries that have been particularly hard hit by the Great Recession, but there also countries such as Germany or Poland that have gotten much better through the crisis. For each country we study four austerity packages that were introduced by the respective governments during the Great Recession. For comparative purposes, we also include an institutional reform proposal for each country in our study. From the perspective adopted here, the austerity proposals (and possibly also the institutional reform proposals) of the governments constitute “proposals at risk,” which are likely to be challenged by some actors mobilizing in the name of aggrieved groups in society. However, not all such proposals have been challenged, nor have all been challenged to the same extent. We shall not only describe how, by whom, and to what extent the different proposals have been challenged but also try to account for the differences in the contentiousness of the challenges and their outcomes.

For this study, we have developed what we call “Contentious Episodes Analysis” (CEA), a novel approach to the study of “contentious episodes” that aims at a more systematic analysis of the dynamics of interaction in such episodes. In this chapter, we shall introduce the broad outlines of our new approach, which is situated in the “middle ground” between the encompassing chronology of the episode, reproduced in narratives, and the micro level of the events, reproduced in simple event counts. Initially, we provide some arguments explaining why we have chosen to study the middle ground. Then we proceed to introduce the conceptual building blocks of our approach. Finally, we provide a brief summary and an overview over the contents of the present volume.

Why and How to Study “the Middle Ground”

In his book on contentious performances, Charles Tilly (Reference Tilly2008: 206) proposed to distinguish between three levels of analysis for studying contentious performances: the reconstruction of single events as one action or interaction after another (what he called the “narrative”), the count of contentious events (what he called “epidemiology”), and the close description of successive interactions within contentious episodes (what he called “the middle ground”). He advocated study at the middle ground, and suggested that from this level, we can move to either one of the other levels but also in a third direction – toward analytic sequences transcending any particular episode but identifying recurrent actions and relations. In this third vein, Tilly himself had aggregated verb categories (e.g. “attack,” “control,” “bargain”) when comparing sets of episodes and then showed which sets of relations among claimants and objects of claims prevailed within different verb categories.

The “narrative” approach is the conventional storytelling of historians, where explanation takes the form of “an unfolding open-ended story fraught with conjunctures and contingency, where what happens, an action, in fact happens because of its order and position in the story” (Griffin Reference Griffin1993: 1099). In contrast, the “epidemiological” approach relies on conventional Protest Event Analysis (PEA). Here, the individual event constitutes the unit of analysis. According to this approach, we can describe an episode in terms of its aggregate event characteristics (e.g. the number of protest events in an episode, the number of events produced by different types of challengers) as well as in terms of the dynamic development of events over time (e.g. the weekly counts of protest events). The “middle ground,” by contrast, focuses on the interactions between challengers and authorities.

Most notably, this middle ground has been the focus of the programmatic Dynamics of Contention (DoC) (McAdam et al. Reference McAdam, Tarrow and Tilly2001). The goals of this seminal study were manifold. Among others, it aimed to (a) overcome the prevailing static approaches in social movement studies, (b) extend the field of study to include other types of actors, (c) introduce a new language to describe/reconstruct processes of contentious politics, and (d) explore the black box between independent and dependent variables, that is, to identify the mechanisms connecting the two. Reflecting on the book’s impact ten years later, its authors (McAdam and Tarrow Reference McAdam and Tarrow2011) self-critically observed that they might have been trying to do too many things, that they had invoked too many mechanisms too casually,Footnote 1 that they had been too indifferent to measurement, and that theirs had still been a state-centric bias. We might add, most importantly, that they failed to provide a framework for the systematic study of interactions across a set of contentious episodes.Footnote 2

Building on DoC, our goal is to further explore the “middle ground.” We do so because we share Tilly’s (Reference Tilly2008: 21) view that this level of analysis offers the “opportunity to look inside contentious performances and discern their dynamics” without losing the opportunity to systematically analyze these dynamics. In other words, we suggest that CEA holds out the promise to go beyond the narrative approach by infusing it with the rigor and explicitness of PEA without losing its dynamic quality. At the same time, CEA aims to move beyond a narrow focus on protest activities by challengers (as in PEA-based research). In this it follows political claims-analysis (Koopmans and Statham Reference Koopmans and Statham1999).

In addressing the middle ground, the challenge is to provide an analytical approach to the study of the dynamics of contention that allows for the systematic comparative analysis of causal patterns across single narratives. Instead of comparing entire narratives (as in sequence analysis), the strategy we propose in CEA is to break down the narratives into their component elements. This is in line with DoC, which insisted on the analysis of smaller-scale causal mechanisms that could then be concatenated into broader processes, that is into “causal chains,” as Gross (Reference Gross2018) has proposed to call such sequences of mechanisms. As Gross points out (Reference Gross2018: 345), such causal chains have structures that may vary in the time period over which a mechanism sequence unfolds, in the number of mechanisms tied together, in the levels of social complexity spanned by actors, and in the abstract patterns formed by connections across mechanisms. CEA is designed to study such sequences of mechanisms systematically.

Generally, the goal of CEA is to specify the concepts of DoC in such a way that they can be applied to systematic comparative analyses across episodes.Footnote 3 Before introducing the building blocks of the proposed CEA in more detail let us highlight three more general points: First, in conceptual terms, CEA privileges the interaction between governments and their challengers. While this focus allows us to move away from the “starkly Ptolemaic view of social movements” that puts movements at the center of the political universe (McAdam and Boudet Reference McAdam and Boudet2012), it keeps the state-centric perspective of DoC and its inherent limitations. It does so by largely drawing on the political process model, which has long since argued that social movements are sustained interactions between challengers and powerholders (Tilly Reference Tilly1978). In this perspective, challengers’ actions can only be understood in relation to the actions by authorities. Relatedly, we shall only allow for a rather reductionist conceptualization of other participants in the episodes, and we shall limit the possible action repertoires of the various actors, too. In other words, there is a price to be paid for the systematic approach we propose here. In a way, the ontology we propose is rather “flat.” That is, in line with the tradition of social movement research inaugurated by Tilly (Reference Tilly1978), CEA adopts a structural-relational perspective focusing on interactions between challengers and authorities, neglecting other components of the mobilization process – in particular, the subjective dimension of contentious politics including processes of framing, the construction of collective identities, emotions, motivations, beliefs, and values. While focusing on interactions, CEA is distinguished from relational accounts of social movements focused on dynamics of interactions within social movements: that is, on interactions among challengers (see Diani Reference Diani2015), thus excluding the analysis of groups, or “catnets” (Tilly Reference Tilly1978). In contrast, CEA aims to build the sequence of interactions within an episode by considering the actions by several types of actors – challengers as well as authorities and third parties – and it proposes a fairly parsimonious conceptualization of the action component on which it focuses.

Second, some details of the suggested approach are tailored to the examples that we shall study here – sixty contentious episodes that have taken place in Europe in the course of the great financial and economic crises that shook the continent from 2008 to 2015. However, we would like to insist that the approach is more flexible than it might seem at first. It is, for example, not restricted to interactions between the government and its challengers in the public arena. The type of arena and the type of actors studied may vary. For example, one might study the interactions between challengers and other types of authority – such as supranational or local political authorities, churches, business corporations, or media – or focus on the interactions between movements and countermovements. What we would suggest, however, is that one cannot do all these possible applications at the same time. In order to keep any analysis manageable, we have to make choices depending on the specific research questions.

As with classical PEA, we think that this flexibility might also be a major strength of the approach (see Beissinger Reference Beissinger2002: 460f.). That is, CEA provides a common conceptual language and general guidelines for data collection and analysis, but ultimately researchers can and should adapt it to the specific research questions at stake. In our study, we ask questions about the variety of contention related to economic and institutional reforms in the Great Recession with regard to the intensity of conflict, the actors involved, the configurations of actor coalitions, and their action repertoires, as well as the outcome of the episodes. In addition, we ask about patterns of interaction in the course of the episodes – interactions between the two main contestants, government and challengers, and between each one of them and potential third parties. We are also interested in identifying critical moments in an episode that decisively redirect the sequences of actions from one state of interaction dynamics to another.

Finally, let us point out that it is possible and, indeed, necessary to complement the bare bones of an analysis based on the CEA framework we propose here with narratives (or process tracing) in order to get a more complete account of the dynamics of contention in the episodes in question. We would maintain that the skeleton of the structural-relational analysis we propose here will make it easier to put flesh on the bare bones in order to get to a full understanding of the episodes one is studying and to systematically compare the various cases. However, it only complements but does not replace a more qualitative and in-depth analysis.

The Conceptual Building Blocks of the Contentious Episode Analysis

Our conceptualization of “contentious episodes” follows the tradition of DoC (McAdam et al. Reference McAdam, Tarrow and Tilly2001), but we have adapted it to our specific purposes. McAdam et al. (Reference McAdam, Tarrow and Tilly2001: 5) defined contentious politics as

episodic, public, collective interaction among makers of claims and their objects when (a) at least one government is a claimant, an object of claims, or a party to the claims and (b) the claims would, if realized, affect the interests of at least one of the claimants.

They defined episodes as “continuous streams of contention including collective claims making that bears on other parties’ interests” (p. 24). More than a decade later, Tilly and Tarrow (2015: 7) reiterated and clarified the notion of contentious politics:

contentious politics involves interactions in which actors make claims bearing on other actors’ interests, leading to coordinated efforts on behalf of shared interests or programs, in which governments are involved as targets, initiators of claims, or third parties. Contentious politics thus brings together three familiar features of social life: contention, collective action, and politics.

For Tilly and Tarrow contentious claims making becomes political when the interaction involves agents of governments. Closely following these conceptions, but simplifying them, we define a contentious episode as a “continuous stream of interactions regarding policy-specific proposals between the government and its challengers, involving also some other actors.” In other words, for us, the key defining element of a contentious episode is the dyadic interaction between two stylized types of actors – the government and its challengers – each making claims on behalf of its own interests and/or on behalf of some other actors.Footnote 4 Consequently, CEA examines the sequence of political claims making by different actors within broader episodes. In doing this, it goes beyond political claims making analysis (as for example proposed by Ruud Koopmans and Paul Statham Reference Koopmans and Statham1999). Political claims making analysis indeed overcomes some problems related to PEA (protest event analysis), to the degree that it includes the investigation of political demands regardless of their form in which they are made and regardless of the nature of the actor (Koopmans et al. Reference Koopmans, Statham, Giugni and Passy2005: 254). However, the innovation of CEA lies in conceiving political claims within a sequence of ordered interactions – something that is missing from claims making analysis used in prior studies – thus enabling the possibility to engage fully in making inferences on causality and disentangle the mechanisms at work in interactions.

In the specific cases we study, the government is initiating the contentious episode by introducing a policy proposal into the public debate. Naturally, in the overall universe of cases, not all episodes need to be initiated by the government but our chosen focus on a subset of such episodes, namely policy episodes, entails that governments are the first movers. The challenger is an actor who opposes this proposal by means of “contentious performances” and other public claims making. A general account of such contentious performances should in theory allow for protest actions on behalf of the government, but in our selected cases of policy episodes this is rather unlikely, and we have in fact identified no such actions to warrant this concern.

In addition to the government and the challengers, we introduce a third set of actors who contribute to the sequence of interactions constituting the episodes – a heterogeneous category of third parties. Under this category, we aggregate all the other participants in the episode who intervene on behalf of either the government or the challengers or who try to mediate between the two without being a member of either of the two camps that oppose each other in the contentious episode. In the example we presented at the beginning of this chapter, the category of third parties includes such diverse actors as the MPs of the governing party, opposition parties, and foreign creditors. The main reason for this simplification is that we are mainly interested in their relationships with the two main protagonists but not in the detailed relations between the various types of third parties.Footnote 5 Also, an important caveat for our definition for third parties above is that their role in a particular episode stays more or less unchanged. If an actor starts out as mediator but later gets directly involved in contentious challenges, we regard them as a challenger actor for the purposes of that particular episode.

Our conceptualization of the main actors in the contentious episode is closely related to the arena concept of Myra Ferree et al. (Reference Ferree, Gamson, Gerhards and Rucht2002). We include all actors who are actively engaged in the conflict in the public sphere but exclude mere bystanders who do not get involved and who constitute the audience for the actors engaged in the public arena. Nor does our framework include the public as a specific actor,Footnote 6 but it does not exclude the addition of the public as a fourth actor depending on the research question one might have. We do not include it for the time being to keep the framework as parsimonious as possible.

Episodes then are composed of single actions by one of the three stylized actors interacting with the other actors. In protest event analysis, each action of the challenger constitutes an event. Instead, we shall reserve the term “action” for individual components of episodes. These actions are typically triggered by previous actions of some other participants in the episode. We call a sequence a series of actions in which each component action is triggered by a previous action. In adopting this terminology, we follow Tilly (Reference Tilly2008: 10), who characterizes episodes very broadly as “bounded sequences of continuous interaction.” The focal properties of sequences are order and convergence (Abbott Reference Abbott1983: 133). Order is crucial “for if the order of the sequence has no effect on its future development, there is no need to worry about sequences at all.” Convergence refers to the end point of a sequence and is one possibility among others for a sequence to end (oscillation or divergence being possible alternatives).

Sequences can be of different length. They may range from short exchanges of verbal statements to a sustained series of interactions between the three types of actors. It is possible that an episode consists of a single sequence of interactions of variable length. More likely, however, an episode is composed of a set of partly overlapping sequences. These sequences may be triggered at different moments during the episode by successive actions of the protagonists. In the particularly complex episode of the midterm adjustment program with which we initiated this chapter, we counted no less than 216 partly overlapping sequences, with an average length of sixty-five actions. While the shortest sequence included only two actions, the longest sequence was composed of no less than 109 consecutive actions. Although the sequences were comparatively long in this episode, their pace was high, meaning one action followed rapidly upon the other. Thus, the average duration of a sequence was only eight days and the maximum duration of a sequence did not exceed three weeks.

As Tilly (Reference Tilly2008: 10) suggests, “cutting the big streams into episodes” will usually allow us to get a better grip on the cause–effect dynamics. The question is, of course, how to cut. As already pointed out, we let an episode start at the moment when a government publicly announces the policy proposal that constitutes the focal point of the episode. This implies that the government is the first mover in our type of crisis episode and the challengers are in a situation where they can only react to the government’s proposal. This situation is radically different from one where the challengers propose a reform, proactively attempt to put it on the agenda, and implement it against the opposition of the government (Walgrave and Vliegenthart Reference Walgrave and Vliegenthart2012). There is nothing in CEA that prevents its application to this different kind of situation, however. Episodes can have different types of starting point, depending on the dynamics to be analyzed.

Episodes may also end in various ways. Our types of episode end in one of two ways. They typically end with the formal adoption of the (possibly modified) proposal by the government. The episode of the midterm adjustment program illustrates this type of ending. Alternatively, if the challenger continues to mobilize after the formal adoption of the proposal, the episode may end when the continuous stream of interactions between the government and its challenger related to the proposal breaks off.Footnote 7

The focus on the actions by different types of actor does not come at the expense of focusing on mechanisms. However, it does clearly narrow down the type of mechanism under scrutiny. In CEA, we focus on what McAdam et al. (Reference McAdam, Tarrow and Tilly2001: 26) call “relational mechanisms,” that is on mechanisms that “alter connections among people, groups, and interpersonal networks.” For example, brokerage – a key mechanism introduced in DoC – is an action of a third party that mediates between two contestants that might change their relationship (into a more cooperative direction). Repression, to take another example, is an action by government that “raises the contender’s cost of collective action” (Tilly Reference Tilly1978: 100) and may thus trigger a violent reaction by challengers, and would be a sign of the further deterioration of the relationship between the perpetrator and the target actor. Radicalization by challengers, finally, is a mechanism that may threaten the government and induce it to step up its repression of the challengers’ actions. In the way we conceive of mechanisms, they correspond to specific sets of interactions that concatenate into processes: that is, longer sequences of interactions.

To be sure, there are mechanisms that are not reducible to actions. In addition to relational mechanisms, DoC also distinguishes between environmental and cognitive mechanisms. We shall also rely on environmental mechanisms. They come into play once we introduce context characteristics that condition the interaction dynamics between the key actors (see Chapter 3). However, we largely neglect cognitive mechanisms. Our approach gives short shrift to the fact that politicians may be influenced by social movements in formulating their policy proposals, that they might anticipate movement protest, or that they might test policy before formal announcements in order to come up with a proposal that will find broad acceptance. As mentioned, our action-centered approach comes at the cost of neglecting beliefs and expectations. CEA is focused only on actions, the interrelationship between actions, and the patterning of these interrelationships. While it is true that a lot of what takes place within contentious episodes has to do with people’s expectations and beliefs (not to mention emotions), our approach strips these considerations out of the equation in a way that the case study method does not. We concede this important point, but we would like to suggest that it is possible to use CEA to reconstruct the rough outlines of a given episode that are then taken as the starting point for a more detailed account of the development of this particular episode. We shall provide an example of this possible extension of our approach in Chapter 12, where we present the Greek episodes in more detail.


According to our conceptualization, the government includes all public authorities linked to the government: that is, the head of government and other members of the cabinet as well as all national public officials. These actors are proposing the policy change. Usually, the political parties of the governing coalition are also part of the government. In some cases, however, a party of the governing coalition may be divided on the proposal and this division may have relevant repercussions on the overall conflict. In these cases, the dissenting voices from the governing party may be coded as third parties or even as part of the challenger coalition. Thus, in the episode of the midterm adjustment, the MPs of PASOK, the governing party, are considered as third parties because of their ambivalence with regard to the government’s proposal.

The challenger includes all actors who oppose the government’s proposal at least partly outside the routine, institutionalized arenas of interest articulation by means of sustained and coordinated collective action, possibly on behalf of other opposing groups. Note that this definition follows the general approach in social movement research and excludes actors who voice their opposition only in routine ways in institutionalized channels, such as national parliaments or tripartite bodies of interest representation (these actors are considered third parties). The challenger in a given episode can be (a) an individual organization – nonmainstream parties (such as a populist parties), mainstream opposition parties, public interest groups (NGOs), unions, social movement organizations (SMOs) – or representatives of such organizations; (b) a social movement: that is a coalition of organizations or dense informal interorganizational networks with a strong identity, or their representatives; or (c) a conflict coalition or alliance: that is, a network of organizations without a strong identity (see Diani and Bison Reference Diani and Bison2004). Empirically, CEA can discern social movement dynamics from coalitional dynamics with no identity, given that the former involves sequences of sustained interactions within an episode (sustained interactions are the basis for the development of strong collective identities), while the latter relies on more contingent and less durable sequences of interactions.

All components of “the challenger” share the opposition to the government’s proposal, but given that the proposal may be a package that includes diverse policy measures, they need not necessarily pursue the same targets, nor may they be part of the same coordinated effort to oppose the proposal. In other words, they form an “objective coalition” against the government (in the sense that they all oppose the same package of proposals) but not necessarily a “subjective” one (since they do not necessarily coordinate their efforts in one and the same collective action).

The challenger opposes the government by striving for the “expansion of conflict” to an ever larger public (Schattschneider 1975): that is, it seeks to politicize the proposal by drawing the public’s attention to the proposal (to render it more salient), by mobilizing public resistance against it (to polarize public opinion on the proposal), and by expanding the number of actors opposed to it (see Hutter et al. Reference Hutter, Grande and Kriesi2016). Public claims making is designed to unleash a public debate, to draw the attention of the public to the grievances of the actors in question, to create controversy where there was none, and to obtain the support of the public for the actors’ concerns. Controversial public debates and support by the general public are expected to open up access and increase the legitimacy of speakers and allies of the challenger with journalists and with decision-makers who tend to closely follow the public debates (Gamson and Meyer Reference Gamson, Meyer, McAdam, McCarthy and Zald1996: 288).

As already observed, third parties include all the other participants in the episode who intervene on behalf of either the government or the challengers or who try to mediate between the two without being a member of any one of the two camps. Just like challengers, third parties can cover a highly diverse group of actors in terms of their institutional characteristics: They can be supranational actors, foreign governments, independent regulatory state agencies, opposition parties, or even government coalition members. In the episode of the midterm adjustment, in addition to the MPs of the governing party, third parties also included opposition parties and supranational actors exerting pressure on the government.

As we alluded to above, we can treat the two adversaries and the third parties as unitary actors, but we can also distinguish the actors within the adversarial coalitions according to their institutional characteristics. The institutional taxonomy of actors will depend on the specific episodes. In our study, we coded a very large number of institutional actor types that we inductively reduced to more limited sets of actors. Chapter 6 will introduce more details in this respect.

Finally, note that this approach can also accommodate actors changing from positions of challengers or third parties to those of authorities and vice versa. An example would be the Democratic Party in the U.S.A. As shown by Heaney and Rojas (Reference Heaney and Rojas2015), when discussing the U.S. anti-war movement after 9/11, the Democratic Party initially allied with the anti-war movement in challenging the George W. Bush presidency in an effort to end the Iraq War: that is, it served as a third-party supporter of the movement. In this phase, leaders of the Democratic Party adopted the movement’s issue and frames and worked together with movement activists to implement the movement’s agenda. However, by 2009, after the presidential elections of 2008 won by Obama, synergies between the Democratic Party and the anti-war movement appeared to have largely vanished. “As the Democrats regained control of government … rather than staying focused on their position on a single issue – such as their opposition to war – many partisans gave greater attention to other callings from the Democratic Party” (Heaney and Rojas Reference Heaney and Rojas2015: 5).

The Action Repertoires

The contentious politics scholars have focused their attention on the action repertoires. As Tilly and Tarrow (2015: 39; emphasis in the original) advise: “We can learn a lot from what activists say or later write about their activities. …We will learn more by examining what activists do during major episodes of contention.” Following this advice, we focus on the action component of the series of interactions constituting the episode. In following this approach, we suggest that the main problem of DoC is not so much the multiplicity of mechanisms it introduced but the fact that it too easily dropped the fine-grained analysis of single actions. As a result, the set of mechanisms it introduced was rather unsystematic. We argue that one needs to first focus on the level of the single action before one can systematically start to combine them into more complex sequences.

We conceptualize the action repertoire separately for each one of the three actors. Importantly, we suggest that the action repertoire of each has two dimensions – a procedural one and a substantive one: that is, an actor can relate to another actor in procedural and in substantive terms. The procedural dimension refers to the relationship between two actors. This can range from conflictive to cooperative. The substantive dimension refers to the substance of the actors’ claims that they address to each other. It can range from support to rejection of the government’s proposal. In our proposed toolbox, each actor type has a specific action repertoire in terms of both dimensions. Table 1.1 presents an overview.

Table 1.1. Detailed action repertoires of the three actor types

GovernmentChallengerThird party
support of proposalstickingcooperationsupport of government
rejection of proposalconcessiondisruptive/non-disruptive actionsupport of challenger
conflictiverepressiondisruptive/non-disruptive actionsupport of government/challenger

In the stylized world of CEA, each actor has three basic options. Once the proposal has been launched and challengers have reacted, the government’s three options are repression, concessions (in procedural and substantive terms), and sticking to its policy. Tilly (Reference Tilly1978) had originally also distinguished between three options (all conceived in procedural terms) – repression, facilitation, and toleration. Toleration was defined as a residual category in-between repression and facilitation that included inaction or disregard of protest. More recently, the term “toleration”has been criticized for its vagueness and value-laden implications (acceptance of the challenge). In her analysis of Egyptian protest against the Mubarak regime, Dina Bishara (Reference Bishara2015) proposed to replace it with the term “ignoring,” which ranges from passive to actively dismissive responses, and suggests that severe forms of ignoring can fuel protest by provoking moral outrage and indignation. Samson Yuen and Edmund Cheng (Reference Yuen and Cheng2017), analyzing the umbrella protest in Hong Kong, introduced yet another concept for government reactions to protest that lie between repression and concession – “attrition.” In this case, toleration is only ostensible, while the government actually “uses a proactive action repertoire to discredit, wear out and increase the costs of protest” (p. 613). The proactive action repertoire in the case of the Chinese authorities in Hong Kong included maintaining elite cohesion, mobilizing countermovements, and using the courts for legal action against the protestors. From our perspective, both depreciating statements about the protestors by the Egyptian authorities and the more indirect proactive repertoire used by the Chinese authorities to counter the protests by challenging movements are all procedural reactions that would qualify as “repression” in the broader sense of the term. As a third category between repression and concession, we propose instead the phrase “sticking to the policy proposal,” which means that the government reaffirms its support of the policy in substantive terms. This is, of course, equivalent to ignoring the protests and may have the effect of attrition as well, but it does not include proactive attempts to undermine the effectiveness of protest by dismissive statements or further-going acts of surveillance and under- cover repression.

The challengers react to the kind of proposals we study in our research. The challengers have first to decide whether they want to react at all. Given that the government proactively pursues policy reform, the challengers are in a situation where they can only react by rejecting the government’s proposal in one way or another: that is, in our episodes we are dealing with threat-induced challenges. As argued by Almeida (Reference Almeida2007: 125) against the background of the Latin American experience, “economic austerity policies (e.g. fuel price hikes, privatization of a public service or utility) that are expected to make popular sectors worse off if implemented are likely to set in motion defensive mobilization that focuses attention on the government and state managers.” To put it differently, the challengers in our episodes constitute “movements of crises” (see Kerbo Reference Kerbo1982) that attempt to fend off threatening policy measures. Threat is the cost that protestors will incur regardless of whether they act or not (Goldstone and Tilly Reference Goldstone, Tilly, Aminzade, Goldstone, McAdam, Perry, Sewell, Tarrow and Tilly2001).

We assume that the challengers act rationally and that they adopt strategies that they expect to have the greatest chances of success at the lowest costs. Conventional politics is usually less costly than protest politics (Cunningham Reference Cunningham2013: 293), but the costs of the threat posed by the government’s proposal may be sufficiently high to induce the challengers to act outside of conventional politics. Once the challengers have decided to act, they have basically three options for action during the episode: They can launch a disruptive challenge, or a nondisruptive challenge (both in procedural and substantive terms), or they can cooperate (in the further course of the episode) with the government. Following the lead of claims making analysis, we extend the kind of communicative acts of challengers that we include in the analysis beyond protest events (see Koopmans and Statham Reference Koopmans and Statham1999, Reference Koopmans, Statham, Koopmans and Statham2010: 54f.): Non-disruptive actions such as verbal claims are included in the challengers’ action repertoire as well. We distinguish nondisruptive from disruptive actions based on the extent to which they are conventional and institutionalized.

Finally, the three strategic choices of third parties include (a) siding (in substantive or procedural terms) with the government; or (b) siding with the challengers; or (c) attempting to mediate (in procedural terms) between the two. As already pointed out, we are not interested in the relationship between the third parties among themselves and we do not code neutral positions of third parties unless they proactively engage in mediation between the opposing actors. In general, we expect that third parties are more likely to react in substantive terms: that is, to participate in the debate on the proposal without entering into a debate about the mobilization process. Third parties that engage in favor of a given camp both in substantive and procedural terms can be considered to be stronger allies than those who engage in substantive terms only.

In the midterm adjustment episode, the predominant government strategy was to stick to its proposal. Given the external pressure, it did not have any other option. Subsidiarily, it attempted to fend off the challengers with repressive measures. The challengers on the other hand insisted that the government should withdraw the proposal, and they mobilized massively in demonstrations and strikes to support their claims. We counted no fewer than forty-three demonstrations and twenty-one strikes during the four-month period covered by the episode. The third parties were about evenly split: Half of their interventions supported the government, while half sided with the challengers.

Building on the action repertoire of the actors involved in an episode, we shall propose a summary measure of the contentiousness of each episode. The three types of actor each contribute to the contentiousness of an episode to the extent that they interact with each other and to the extent that their actions are disruptive (challengers), repressive (government), or one-sided (third parties).


The different types of action constitute the building blocks for the construction of the sequences within an episode. Recall that we define a sequence as a series of actions in which each component is triggered by a previous action: that is, the actions in a sequence are explicitly linked to each other. Sequences have properties of their own that can be studied in descriptive and explanatory terms. Sequence analysis occupies a well-established place among social science methodologies. The most prominent approach to study sequences rests on some assumptions that CEA does not fulfill, however. Moreover, this approach takes the whole sequence as the unit of analysis and attempts to find clusters of similar sequences, a goal that CEA does not try to pursue (see Chapter 6 for more details). Instead, CEA is interested in the overall sequence structure of an episode and in the dynamics of the sequences across episodes.

We shall analyze the overall sequence structure of entire episodes in temporal terms (referring to the duration and pace of the sequences), and in relational terms (referring to their length, breadth, and overall complexity), allowing us to get a first idea of the basic structural properties of the episodes and their determinants. In addition, the reconstruction of the sequence structure also allows us to identify specific important points in the development of the episodes, such as turning points. Turning points can be defined in two different ways. An action may be a turning point in the sense that it leads to a certain closure of the interaction process by closing alternatives and focusing the interaction on a single thread. Both Abbott (Reference Abbot2001) and Ermakoff (Reference Ermakoff2015) define turning points as closing points, but an action may also be a turning point in the sense that it opens up the interaction process by giving rise to a multiplicity of reactions of some consequence – that is, reactions each of which in turn trigger a series of further actions.

Based on the chronological order of the sequences, we shall analyze their interaction dynamics with the aim of uncovering general mechanisms that characterize the episodes triggered by austerity policies in the Great Recession. The gist of the interaction dynamics within an episode lies in the interdependence of the three types of actor. As Beissinger (Reference Beissinger2011: 27) has observed, “one of the defining features of mobilization – and its greatest challenge for causal explanation – is the high degree of inter-dependence of the actions and reactions involved, both within and across episodes of mobilization. While not a feature characteristic of mobilization alone, it figures so centrally in contentious politics that it is difficult to explain any protest episode without fundamentally addressing this issue.” Our approach assumes that the actors involved act retrospectively: that is, they react to the actions of the other actors (see Moore Reference Moore2000: 121).

The most elementary sequence is a pair of consecutive actions. We shall focus on pairs of actions that are chronologically following upon each other within a sequence, one being the trigger of the other. For example, we shall study the reactions of the government to disruptive actions of the challengers. Even with only three types of actor and a limited action repertoire of three types of action per actor, there are multiple patterns of possible interactions in any such pair of actions. In the example, the government can react in three ways to the disruptive challenger action, and one of the possible three pairs would be disruptive challenger action followed by government repression. Importantly, we shall generalize the approach based on pairs in two ways. First, we shall relax the restriction that the action triggering a given reaction must immediately precede the reaction in question: That is,. we shall allow for actions that are chronologically preceding the reaction in question by variable steps in the chain of the action sequence to have an impact on the reaction in question as well. In the example of government reactions to disruptive actions by challengers, we shall study the government’s reaction to immediate challenger actions but also to such actions that are further removed in the sequence. Second, instead of studying specific pairs, for example a disruptive challenger action followed by government repression, we shall include in the analysis of the given type of reaction (for example, government repression) any possible trigger (for example, actions by third parties supporting the challengers, disruptive, and non-disruptive actions by challengers) of the reaction in question. In other words, we shall introduce the different action types of the three stylized actors into the multivariate analysis to explain a given reaction at one and the same time. These extensions will allow us to come up with a more detailed account of the interaction patterns in the various episodes.

Conclusion and Overview over the Volume

In this introductory chapter, we have introduced a set of concepts and general guidelines of what we call Contentious Episode Analysis (CEA). In the footsteps of Dynamics of Contention (DoC), we are attempting to develop a conceptual framework that improves upon the concepts originally introduced by McAdam et al. (Reference McAdam, Tarrow and Tilly2001) and that allows us to study contentious episodes more systematically, in a nonnarrative mode. Our analytical strategy is similar to that of DoC: We also propose to decompose the episodes into their component elements – actors, actions, sequences of actions, pairs of actions – that can then be recombined in a systematic way to account for specific processes in the dynamics of contention. We suggest that CEA holds out the promise to go beyond the narrative approach by infusing it with the rigor and explicitness of PEA, without losing its dynamic quality. At the same time, following Koopmans and Statham’s (Reference Koopmans and Statham1999) claim analysis, CEA aims to move beyond a narrow focus on protest activities by challengers (as typically done in PEA-based research) by incorporating into the analysis a broader set of action repertoires by a broader set of actors (as is typically done in claims analysis). In addressing the middle ground favored by Charles Tilly, we apply an analytical approach to the study of the dynamics of contention that allows for the systematic comparative analysis of causal patterns across individual narratives. We hope that the toolkit we introduce here will allow for a more systematic analysis of a wide variety of questions linked to the DoC.

In the subsequent chapters of this volume we shall elaborate these concepts in more detail and show how they can be operationalized and implemented in the analysis of specific questions. The volume is divided into three parts. In the remainder of the first part, we shall first present the methods we used to collect our data as well as the context conditions of the sixty episodes we study in this volume. In Chapter 2, we set out how we selected the sixty episodes and how we documented them. As for their selection, it is important to note that, following the lead of McAdam and Boudet (Reference McAdam and Boudet2012), we tried not to select on the dependent variable: Our selection procedure is not based on whether or not there was a serious challenge to the government’s proposal. The documentation of the episodes involved the selection of articles in national quality newspapers and the manual coding of these articles. Chapter 3 puts our sixty episodes into their economic and political contexts. It clarifies that the actual decline in economic performance was much more strongly and sharply felt in the south than in the other two parts of Europe. Moreover, it shows that the governments in the hard-hit countries got under double economic and political pressure. In terms of the timing of the proposals, this chapter finds that it has been closely related to both the development of the economic crises and to strategic political considerations.

Part II elaborates the various key concepts, introduces their operationalization, and presents results at the level of the episodes. It provides an overview over the varieties of contention that we observed during the Great Recession. Building on the action repertoires, Chapter 4 introduces the multi-dimensional concept of contentiousness, describes the contentiousness of the individual episodes in the different countries, and provides a set of factors that contribute to the episodes’ contentiousness. As it turns out, the Greek episodes (among them the midterm adjustment program that we introduced in this introductory chapter) have been the most contentious of all. In contrast, the German episodes were the least contentious ones. Of course, that is not so surprising given that Germany got through the Great Recession better than any of the other countries in our study. Chapter 5 presents the actors who have been involved in the various episodes, and characterizes them in institutional terms. Among other things, we find that labor unions have been the most important challengers during the contentious episodes in the Great Recession. The chapter also analyzes the actor coalitions and configurations in the various episodes. Chapter 6 introduces the analysis of the sequences. It characterizes the episodes according to the temporal and relational structure of the sequences. The chapter describes the overall sequence structure of the various episodes and makes an attempt to explain it. Thus, the episode types introduced in Chapter 3 turn out to be the best predictor of the temporal sequence structure, with structural and institutional reforms being characterized by a slower pace than I.M.F. bailouts, bank bailouts, and fiscal measures. The greater pressure associated with the latter episodes leads to a more intensive pace of interaction between the government and its challengers. Chapter 7 concludes Part II with an analysis of the outcomes of the episodes. The results of this chapter show that there was very little government responsiveness to challenger actions. Only exceptionally, in the case of episodes proposing extremely severe measures, did governments make some limited concessions to the challengers.

Part III presents various aspects of the dynamic interactions during the episodes. Chapter 8 sets the stage. It introduces the specific method we apply to studying the dynamic interactions between the three actor types, and it tests some general hypotheses concerning their interactions. For all action forms of both adversaries, it finds strong evidence for path dependence, with the pattern being somewhat stronger on the government side. By and large, government behavior appears to be independent of previous challenger actions. With respect to the impact of third parties, governments have a higher propensity to repress challengers when they are not supported by third parties. Most importantly, however, governments seem to honor mediation attempts with concessions. The analyses in this chapter do not take into account the context of the various episodes, however. It is the following two chapters that introduce context into the analysis: Chapter 9 focuses on government reactions to challengers, while Chapter 10 deals with the challengers’ reaction to government repression. Both chapters indicate that context is very important. The results of these chapters are rather complex, and they tend to qualify the sweeping results of Chapter 8. Thus, the mediation effect which was uncovered in Chapter 8 appears to be limited to the least threatening episodes – the party-driven episodes, and even in these instances it has at best been marginal. In contrast, the intervention of international actors on behalf of either the government or the challengers tends to be more consequential, especially in party and movement-driven conflicts.

The last two empirical chapters adopt a somewhat different approach. Chapter 11 analyzes the two types of turning points in more detail and uses this concept to distinguish between different phases of the episodes – the opening phase, the main phase, and the closing phase. It shows that the government is mostly responsible for the turning points, and that it dominates in the opening phase, while the challengers play a much bigger role in the main phase and, above all, in the closing phase. There are signs of escalation in the closing phase of the episodes. Chapter 12, finally, shifts gears once again and shows how CEA can be used in a more qualitative way to analyze a series of episodes that have taken place in one country. The case studied in this chapter is Greece, the country that stands out for the extreme contentiousness of the episodes unleashed by austerity packages during the Great Recession. It treats the contentious episodes of this country not as separate units of analysis, as do the rest of the chapters, but as parts of a larger campaign that unfolded during the years Greek politics was dominated by the bailout. This chapter uses the contentious episodes as a guide to build a narrative account of Greek contention during the age of the bailout.

The final chapter 13, concludes. It draws together the various threads of the empirical chapters and presents our own assessment of the novel approach for the study of political contention that we introduce with this volume.


Parts of this chapter are taken from: Hanspeter Kriesi, Abel Bojar and Swen Hutter. 2019. “Contentious Episode Analysis”, Mobilization 24 (3).

1 Lichbach (Reference Lichbach, Davenport, Johnston and Muelle2005: 228) has already pointed to the key problem of the introduction of mechanisms – multiplicity. As an antidote he proposed to embed causal mechanisms in theories and evaluate the mechanism by using stylized facts and historical narratives. He argued that generating mechanisms is easy but locating them is not. In his view, the real challenge is to embed mechanisms in larger and more organized structures of knowledge so as to deepen our understanding of interesting and important causal processes (Lichbach Reference Lichbach, Davenport, Johnston and Muelle2005: 233f).

2 The same applies to Alimi et al. (Reference Alimi, Bosi and Demetriou2012), who adopt the mechanism approach but who use it in their narratives in a rather loose sense.

3 In this respect, our approach differs from the one chosen by Griffin (Reference Griffin1993), who relied on “event structure analysis,” a procedure developed by Heise (Reference Heise1989) that allows reconstructing the causal structure of the narrative about an individual episode – in his case a lynching episode that took place in Mississippi in 1930.

4 Similarly, Ermakoff (Reference Ermakoff2015: 96f.) distinguishes between two general “stakeholders” – challengers and target actors (the target of the challenge), while Biggs (Reference Biggs2002) distinguishes between labor and capital, or workers and employers.

5 In the real world, the boundary between the two camps and third parties is sometimes difficult to draw because members of either camp may distance themselves from their home turf during the episode-specific controversy, trying to operate rather as a third party.

6 In specific instances, the general public is included in the analysis when there is a particular action that is attributable to it, with the referendum on Brexit being a prominent example.

7 As a rule of thumb, we use a period of two months for assessing whether the interactions related to the proposal have indeed come to a halt before we can declare the episode to have ended.

Figure 0

Table 1.1. Detailed action repertoires of the three actor types

You have Access

Save book to Kindle

To save this book to your Kindle, first ensure is added to your Approved Personal Document E-mail List under your Personal Document Settings on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part of your Kindle email address below. Find out more about saving to your Kindle.

Note you can select to save to either the or variations. ‘’ emails are free but can only be saved to your device when it is connected to wi-fi. ‘’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.

Find out more about the Kindle Personal Document Service.

Available formats

Save book to Dropbox

To save content items to your account, please confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account. Find out more about saving content to Dropbox.

Available formats

Save book to Google Drive

To save content items to your account, please confirm that you agree to abide by our usage policies. If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account. Find out more about saving content to Google Drive.

Available formats