Promoting or Preventing Social Change

Jacquelien van Stekelenburg

We organize a workshop, titled "Political Protest and Social Movements" at the annual meeting of the Dutch/Flemish Political Science Associations (the "Politicologenetmaal"). This will take place Thursday, June 7, 2018, at Leiden University. We aim to bring together scholars working on political protest and social movements (see Call for Papers below) Deadline for paper proposals is March 15, 2018. Proposals should include an abstract (max. 250 words), name, affiliation and contact email. You will hear from the workshop coordinators by 31 March 2018, whether your proposal was accepted.

With the Netherlands having the world's highest cycling rates, many see its current cycling policies as a model for the transition to sustainable transportation systems. Understanding these high cycling rates, however, requires understanding the geographic, historical, and institutional context in which social movements, working with government actors, helped stop a rapid decline in cycling rates between the 1950s and the 1970s in the Netherlands. This article uses historical sources and interviews with activists and government actors to show how social movements supported cycling by helping reverse the negative effects of rapid motorization. These social movements worked with government actors to implement three specific innovations: the woonerf, the bottleneck memorandum, and the car-restricted city center. This article contributes to transitions literature by looking beyond the relationship between enterprise and the state and demonstrating how social movements within a specific institutional context and with broad public support can advance sustainable transportation innovations.

social movements and the dynamics of collective action Jacquelien van Stekelenburg and Teodora Gaidytė This chapter is an update of the 2013 chapter Social movements and the dynamics of collective action (Klandermans & Van Stekelenburg, 2013). In that chapter we described fundamental social psychological processes-social identity, social cognition, emotions and motivation-as they were employed in the context of social movement participation and as we argued, engaging in social movements most of the time implies taking part in some form of collective action. However, the world of protest has changed profoundly since 2013. Virtualization gave the world a new 'virtual look' and we observe as much protest now as in the roaring '60s, a decade characterized by high levels of protest activity. It is therefore not surprising that research on the social and political psychology of protest has expanded since 2013, both theoretically and empirically. This chapter is an attempt to synthesize these recent efforts and revise the assessment of where we are. To do so, we refreshed the whole chapter.

PROMOTING OR PREVENTING SOCIAL CHANGE Instrumentality, identity, ideology and group-based anger as motives of protest participation Jacquelien van Stekelenburg Amsterdam, 2006 VRIJE UNIVERSITEIT PROMOTING OR PREVENTING SOCIAL CHANGE Instrumentality, identity, ideology and group-based anger as motives of protest participation ACADEMISCH PROEFSCHRIFT ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam, op gezag van de rector magnificus prof.dr. T. Sminia, in het openbaar te verdedigen ten overstaan van de promotiecommissie van de faculteit der Psychologie en Pedagogiek op dinsdag 23 mei 2006 om 13.45 uur in de aula van de universiteit, De Boelelaan 1105 door Jacquelien van Stekelenburg geboren te Purmerend promotor: prof.dr. P.G. Klandermans copromotor: dr. W.W. van Dijk To the three most important men in my life About being grateful The work with this dissertation has been exciting, instructive, and above all fun. Without help, support, and encouragement from several people, I would never have been able to finish this work. First of all, I am grateful to my supervisor, Bert Klandermans. Bert, I’d like to thank you for your loyalty and integrity. I am grateful you gave me the freedom to do it my way and I admire the way you were able to keep your sense of humor even when the going was going very tough. Above all, you taught me what science is about, not only how to conduct research and writing it down but you gave me a ‘fatherly’ introduction into the science community with its own culture, rituals, and jargon making it a less ‘far from my bed’ experience. I will also give a special thanks to Wilco, my co-advisor. Wilco, I respect how fast you hooked up in the project and were able to provide feedback, thanks! It has been, and still is, a great pleasure to work at the department of SCW of the Vrije Universiteit. I wish to express my thanks to other members of the department. A special thanks goes to the people from ‘Social Conflict and Change’. Not only were they always there for constructive discussions about research related topics, the social atmosphere was also very good (en hij gelooft er in, hij ziet toekomst in de afdeling…). I am very grateful to Conny and Annette, of course we had discussions about social movements, motivation, extreme right movement, women’s movement….but there was so much more to talk about! It is just a tremendous lot of fun to analyze the cultural processes of the science community and the position of women within this community. I feel privileged to have the two of you as ‘personal coaches’. Then there were the little ‘clubs’, such as intervisie every 6 weeks since 1995. The last 10 years our little group experienced all the ups and downs of life in a nutshell, it’s great to know that the intervisie group will always support you, writing application letters, problems on the work floor, problems in the relational atmosphere you name it, the intervisie group is there to support you. Ladies it is been a pleasure, and I am looking forward to the next 10 years! I am also thankful to another club, the AiO-overleg of the department of Social Psychology of the Vrije Universiteit. After changing the Social Psychology department for the Social Cultural Department, the AiO-overleg of social psychology provided a home for me to stay in touch with social psychology. A special thanks goes to Nickie, Myrke, Margreet, Petra, Tom, Nils, Camiel, Petra, Lotte, Daniel, and Frederique. You all made me feel welcome, despite the fact that I was a deserter. I am also very grateful to the ladies of the fightclub. Although we knew each other already for years the death of Mirjam brought us together and gave me the energy to fight for fun in life! All the dinners we have together are enormously valuable to me and give me the strength to cope with all the unfairnesses in life. Annemieke, Annette, Bianca, and Marjoka let’s keep on fighting! The last club I’d like to thank is the RFT-reading club. It is a pleasure to realize that other people find some theories complex too, and it is even more fun to discuss articles and our own work with you. Thanks for all the feedback and interesting discussions. They gave me energy and the idea that science is so much fun. Guda, Krispijn, Belle en Collette, thank you. Friends are a great gift in life, during ski-holidays, summer-holidays, weekends, parties, get-togethers, dinners you were all there, to show me that there is so much more in life than writing a dissertation. Cynthia, Wendy, Marieke, Max, Wim, Sioe, Eelco, Mark, Madeline, Astrid, Hans, Sandra, Esther, Gerard, Paul, Toby and Cristien, thanks! My parents, and my sisters receive my deepest gratitude and love for their dedication and the years of support during my undergraduate studies that provided the foundation for this project. My father for his interest and discussions about the world around us and my mother for teaching me what motivation and perseverance is about. My sister Mariska for loving me and being proud of me ‘at long distance’ and my sister Annoesjka for just having fun about nothing, and organizing chill weekend-stays in Sint Maartenszee. Finally, I want to thank the three most important men in my life. I’d like to thank Mika en Biko for being such a sweet distraction. The two of you gave me the opportunity to put writing a dissertation into perspective. However, most importantly, none of this would have been possible without the love, trust and patience of Pascal. You were the one who encouraged me to take up this project with the words: people should always pursue their dreams. Hereby I express my heart-felt gratitude to you and now it is your time to pursue your dreams. CONTENTS Chapter 1. Introduction and theoretical background. ................................................................... 1 1.1 Introduction........................................................................................................................ 1 1.2 Theoretical background: Social psychological approaches to protest participation. ........ 5 1.2.1 Instrumentality.............................................................................................................. 6 1.2.2 Identity.......................................................................................................................... 9 1.2.3 Group-based anger ..................................................................................................... 14 1.3 Introduction of a new social psychological approach to protest participation. ............... 17 1.3.1 Instrumental path to protest participation................................................................... 19 1.3.2 Identity path to protest participation. ......................................................................... 20 1.3.3 Group-based anger path to protest participation. ....................................................... 20 1.3.4 Ideology path to protest participation......................................................................... 20 1.3.5 An integrative model of the instrumental, identity, ideology and group-based anger paths........................................................................................................................... 21 1.4 Regulatory Focus Theory ................................................................................................ 27 1.4.1 Distinct goals, needs and values................................................................................. 29 1.4.2 Regulatory focus as moderator of the paths taken to protest participation. ............... 30 1.4.3 Regulatory fit: social movement context and regulatory focus.................................. 31 1.5 This study......................................................................................................................... 32 Chapter 2. Development and validation of RFQ-proverb. ......................................................... 34 2.1 Available assessment techniques to measure chronic regulatory focus. ......................... 34 2.1.1 Selves Questionnaires ................................................................................................ 34 2.1.2 Strength-of-Guide measure ........................................................................................ 36 2.1.3 Regulatory Focus Questionnaire by Higgins et al...................................................... 37 2.1.4 Regulatory Focus Questionnaire by Lockwood et al. ................................................ 37 2.1.5 Possible limitations of the RFQs of Higgins et al. and Lockwood et al. ................... 38 2.2 Proverbs as elements of an RFQ...................................................................................... 39 2.3 Development and evaluation of RFQ-proverb. ............................................................... 40 2.3.1 Scale development...................................................................................................... 40 2.3.2 Scale characteristics ................................................................................................... 42 2.3.3 A short version ........................................................................................................... 46 2.3.4 Relationships with other scales .................................................................................. 46 2.4 Study 1 and Study 2 ......................................................................................................... 49 2.4.1 Study 1 and Study 2 Methodology. ............................................................................ 50 2.4.2 Results and Discussion Study 1 and Study 2. ............................................................ 50 2.5 Study 3 ............................................................................................................................. 53 2.5.1 Methodology Study 3. ................................................................................................ 53 2.5.2 Results and Discussion Study 3. ................................................................................ 54 2.6 Conclusions...................................................................................................................... 57 Chapter 3. Methodology............................................................................................................. 60 3.1 Comparative research designs within protest participation research............................... 62 CONTENTS 3.3 Protest surveys. ................................................................................................................ 64 3.4 Procedure Study 4 and Study 5........................................................................................ 66 3.5 Participants Study 4 and Study 5. .................................................................................... 67 3.6 Representativeness of the samples of Study 4 and Study 5............................................. 69 Chapter 4. Instrumental, identity, ideological, and group-based anger and context: the relative weight of participation motives. ................................................................................................. 72 4.1 Method ............................................................................................................................. 73 4.1.1 Sample ........................................................................................................................ 73 4.1.2 Measures..................................................................................................................... 73 4.2 Results.............................................................................................................................. 74 4.2.1 Preliminary analyses................................................................................................... 74 4.2.2 Analyses strategy........................................................................................................ 75 4.2.3 Main effects................................................................................................................ 77 4.2.4 Moderating effects of identity and social movement context. ................................... 79 4.2.5 Mediating effects of instrumental and ideology motives. .......................................... 84 4.2.6 Structural equation modelling. ................................................................................... 87 4.3. Conclusions................................................................................................................... 101 4.3.1 Model accounting for the strength of motivation to participate in political protest. 101 4.3.2 The influence of identification processes on protest participation........................... 102 4.3.3 The influence of social movement context on motivational patterns....................... 104 Chapter 5. Regulatory focus and political protest participation............................................... 107 5.1 Regulatory fit: social movement context and regulatory focus. .................................... 108 5.1.1 Group attractivity and regulatory focus. .................................................................. 108 5.1.2 Collective action frames and regulatory focus. ........................................................ 109 5.1.3 Frame alignment and regulatory focus..................................................................... 111 5.1.4 Determining regulatory characteristics of the collective action frames................... 112 5.1.5 Hypotheses about regulatory fit and participation motives...................................... 114 5.2 Measures ........................................................................................................................ 115 5.3 Results............................................................................................................................ 116 5.3.1 Preliminary analyses................................................................................................. 116 5.3.2. Main analyses .......................................................................................................... 118 5.3.3 Does inner obligation mediate between prevention-focus and identity motives?.... 120 5.4 Discussion...................................................................................................................... 123 5.5 Does regulatory focus moderate the paths taken to protest participation? .................... 124 5.5.1 Regulatory focus and participation motives............................................................. 124 5.6 Measures ........................................................................................................................ 126 5.7 Results............................................................................................................................ 126 5.7.1 SEM: Baseline model of protest participation steered by regulatory focus............. 126 5.7.2 Multiple-group analysis: regulatory focus and context............................................ 128 5.7.3 SEM: modified model of participation in prevention-oriented protest events......... 132 5.7.4 SEM: modified model of participation in promotion-oriented protest events. ........ 133 CONTENTS 5.8 Discussion...................................................................................................................... 135 Chapter 6. General discussion and conclusions. ...................................................................... 140 6.1 An integrative model accounting for protest participation. ........................................... 140 6.2 The influence of social movement context on the relative weight of the paths. ........... 143 6.3 Regulatory organization-participant fit. ........................................................................ 144 6.4 Regulatory focus as steering mechanism....................................................................... 146 6.5 Protest participation and identity................................................................................... 147 6.6 Protest participation and self-regulatory strategies. ...................................................... 149 6.7 Possible shortcomings. .................................................................................................. 151 6.8 Protest participation and directions for future research................................................. 152 6.8.1 From a static to a dynamic model. ........................................................................... 152 6.8.2 Effects of regulatory focus on protest participation. ................................................ 153 6.8.3 Effects of a variety of emotions on protest participation. ........................................ 154 6.8.4 Effects of political, economic and cultural contexts on protest participation. ......... 155 6.9 Concluding remarks....................................................................................................... 157 Samenvatting ............................................................................................................................ 159 References ................................................................................................................................ 166 1 Chapter 1. Introduction and theoretical background. 1.1 Introduction The general picture of Dutch society was one of steady progress up to 2001. The Dutch were happy and satisfied people (Social and Cultural Planning Office, 2004). However, a break in this trend has occurred. “Since 2001 the Dutch social and political climate has been characterized by unrest. On top of this, the economy has deteriorated. A number of notable events took place, including acts of international terrorism and a political assassination within The Netherlands, with serious political consequences. Although not all of these developments and incidents have a bearing on people’s view on government policy, it is plausible to imagine that a climate of social unrest increases people’s demands on the government, because the population seeks greater protection against perceived risks” (Social and Cultural Planning Office, 2004, p. 71). Political protest is one way to address demands to the government. And indeed, after a relatively quiet period, we saw an increase in willingness to protest in the Dutch population aged 18 years and over (from 51% in 2000 to 55% in 2002, Social and Cultural Planning Office, 2004, p. 45) and an increase in actual protest participation (9% in 2000 to 14% in 2002). These figures seem to indicate that both willingness to participate and actual participation in political protest have increased1. As Demeester (2004 p. 1) states “The Dutch have woken up from a long winter sleep”. On Saturday 2 October 2004 more than 300,000 protesters took to the streets in Amsterdam. Most of these protesters (about 250,000) had been mobilized by three trade union federations. This was the biggest trade union demonstration in Dutch history and the biggest demonstration of any kind in Amsterdam since the anti-cruise missiles demonstration in 1981 (400,000 protesters). Why would someone decide to go to Amsterdam and participate in a mass demonstration? Indeed, why would someone participate in a political protest event? Over the last two decades, social psychologists have begun to investigate individual participation in episodes of collective action and political protest. They began by demonstrating that instrumental reasoning controlled people’s participation in collective action. Gradually, they explored other motives that stimulate 1 This dissertation employs several terms for political protest participation, such as: participation in collective action, (political) protest behaviour and participation in unconventional political action. All these terms imply a group member engaging in collective action any time he or she is acting as a representative of that group and the action is directed at improving the conditions of the entire group (Wright, Taylor & Moghaddam, 1990). CHAPTER 1 2 people to engage in collective action. The first motive to be added was identification and, recently, group-based anger has been put forward as another motive. We have now reached the point where, from a social psychological perspective, three possible motives lead to participation in political protest: instrumentality, identity and groupbased anger. This state of affairs leaves many issues unexplored. Klandermans (2003) points to a set of unanswered questions about protest participation. The first unanswered question concerns the role of ideology and its relationship to emotions. In his words: “Strangely enough, very little systematic empirical work is available on ideology and on the way people’s ideals and values generate passionate politics” (p. 699). Indeed, Klandermans makes a plea for a fourth motive that may lead to protest participation: ideology. Another unanswered question is about the relative weight of instrumentality, identity, ideology and group-based anger. Again in his words: “Each mechanism has its impact, but how do they work together? Do they add to each other, or do they interact? Are they correlated or independent determinants of participation?” (p. 699). Moreover, he makes a case for the study of identity in systematic empirical work. As he argues: “A rapidly growing literature exists of identity and movement participation, but most of this literature is only conceptual and very little of it is empirical” (p. 699). Lastly, he points to a set of unanswered questions concerning the dynamics between the individual protester and mobilization strategies of movement organizations: “It would be interesting to know whether a specific strategy of consensus mobilization would activate a specific group of people” (p. 699). The main goal of this dissertation is to develop a theoretical model of political protest participation that speaks to these questions. Our theoretical model is designed to amend and elaborate on the motives currently accepted by social psychologists for participation in political protest: instrumentality, identity and group-based anger. Indeed, our first objective is to integrate the three motives in a single model. We do not believe, however, that the three motives cover the whole range of fundamental motives that make people tick. Therefore, our second objective is to extend the theoretical model of political protest participation by adding an ideology motive. Thus, we propose a social psychological approach to protest participation consisting of the following four motives: instrumentality, identity, ideology and group-based anger. Our third objective is to account for the influence of social movement context on the relative weight of the four participation motives. The current social psychological literature on protest participation does not elaborate on which of the three motives proposed so far will prevail for whom, when, and why. In other words, for whom will what pathway to collective action prevail, and why? The answer to these kinds of questions presupposes a steering mechanism. This brings us to our fourth objective: in INTRODUCTION AND THEORETICAL BACKGROUND 3 addition to the four possible paths to collective action we conceive of a steering mechanism that can explain why one path to protest participation will prevail over another. Regulatory focus is the steering mechanism we propose. This dissertation has at its core the idea that people’s motivation to participate in political protest stems from self-regulation mechanisms originating in a perceived threat to their needs, goals and values. Regulatory focus theory is a motivation theory based on the self-regulation principle (Higgins, 1997; 1998). This self-regulation mechanism can be either promotionfocused, striving for nurturance needs, ideal goals (such as hopes and wishes) and self-directed and autonomous values, or prevention-focused, aiming at security needs, ought goals (such as obligations and responsibilities) and traditional and conformity values. Thus, the two foci are the source of specific goals, needs and values and therefore provide different answers to the question “What is my relation to the world?” Strauman (1996) refers to this question by describing a regulatory focus as a worldview—a tendency to construe situations and experiences in terms of the core psychological situation represented by the two foci. Regulatory focus can thus be seen as a general worldview and has cognitive, emotional and behavioural consequences for how individuals respond to the world (Higgins, 1998). We maintain that regulatory focus influences a frame to interpret the social and political world, and therefore that it shapes protesters’ participation motives. Finally, our fifth objective is to investigate in what social movement context which motives will prevail. We investigate whether regulatory focus influences which collective action frame will be persuasive for whom. Do organizations framing a protest activity in promotion terms attract more promotion-focused than prevention-focused protesters? Or, alternatively, do organizations framing a protest activity in prevention terms attract more prevention-focused than promotion-focused protesters? This relates to the dynamics between the individual protester and mobilization strategies of movement organizations, in other words person-environment fit or, in Higgins’ (1997; 1998) words, regulatory fit. Overview of this book. The thread running through this dissertation is the relation between regulatory focus, participation motives (instrumental, identity, ideology and group-based anger) and social movement context. We thus bring together motivational theory and theories on protest behaviour. This chapter therefore starts with a discussion of the two literatures. Chapter 2 deals with the construction and validation of an instrument to measure regulatory focus, and discusses three CHAPTER 1 4 studies conducted to test the reliability, validity and predictive value of the instrument constructed. The instrument thus tested was employed in all the other studies reported. The third chapter describes procedures, research methods used and data collection methods. It should be noted that a description of the methodology is not confined exclusively to this chapter. Each of the subsequent empirical chapters has its own methodology section in which more specific information regarding operationalization and measures are presented. The findings reported in Chapters 4 and 5 are based on the field study described in Chapter 3. This study was conducted during a demonstration organized by two different movements with two different action orientations (Turner & Killian, 1987): a more poweroriented action against austerity plans regarding early retirement rights and a more value-oriented action against neo-liberal politics. This demonstration offered a unique opportunity to test the relative weight of the participation motives (Chapter 4) and the relation between regulatory focus and the motivational pattern of the protesters (Chapter 5). Chapter 4 reviews literature on motivations to participate in political protest and the influence of the mobilizing context. Subsequently, the results will be presented wherein the relative weight of the participation motives—instrumentality, identity, ideology, and group-based anger—in the two mobilizing contexts will be assessed. Because the action orientations of the two demonstrations differ (one more value-oriented and one more power-oriented) comparing the relative weight of the participation motives offers an opportunity to see whether the relative weight of the participation motives changes as the action orientation varies. Chapter 5 discusses the explanatory potential of regulatory focus theory regarding protest participation. First, we review the literature on collective action frames and present results testing the idea that a collective action frame defined in prevention terms is more persuasive for prevention-focused individuals, whereas a collective action frame defined in promotion terms is more persuasive for promotion-focused individuals (i.e., there is regulatory fit). Thus, we explore whether regulatory focus influences which collective action frame will be persuasive for whom. In our research we exploited the fact that two movements were proposing a different appraisal of the situation. This provided the opportunity to test the idea, articulated as the fifth objective of this dissertation, that organizations framing a protest activity in promotion terms attract more promotion-focused than prevention-focused protesters, whereas protest activities framed in prevention terms will attract more prevention-focused than promotion-focused protesters. In the second part of Chapter 5 we test the extent to which regulatory focus generates specific participation motives. As indicated, we conceive of regulatory focus as a steering mechanism that may explain which pathway to collective action will prevail for whom, and why. INTRODUCTION AND THEORETICAL BACKGROUND 5 We assume that regulatory focus influences the frame to interpret the social and political world, and therefore shapes protesters’ participation motives. Together, Chapters 4 and 5 aim to investigate the unexplored issues mentioned in this introduction. In both field studies, “real” protesters are the subject of investigation. The data presented concern their perceptions, motives and feelings in “the heat of battle”. They describe the results of the interaction between protesters and collective action frames of organizations, how collective action frames appeal to people, and regulatory focus and participation motives. In Chapter 6, finally, the reported findings are reviewed and related to the literature, practical implications are considered and recommendations for future research are formulated. 1.2 Theoretical background: Social psychological approaches to protest participation2. Why would someone become engaged in a protest campaign? This question brings us to the level of analysis of the individual and therefore to the realm of social psychology. As social psychology explores the causes of the thoughts, feelings and actions of human beings, and primarily the influence of other people upon these, it has a lot to offer to the study of social movements in general, and protest behaviour in particular. We illustrate this point with the presentation of three social psychological approaches to social movement participation. Over the last two decades, social psychologists have begun to investigate individual participation in episodes of collective action and political protest. Gradually, they have explored more and more motives that stimulate people to engage in collective action. Initially the focus was on the perceived costs and benefits of participation. Indeed, it was demonstrated that instrumental reasoning controlled people’s participation in collective action (Klandermans, 1984). Participation was seen as an opportunity to change a state of affairs at affordable costs. It also became clear, however, that instrumental reasoning is certainly not a sufficient reason to participate in collective action. Gradually, the significance of collective identity as a protest participation motive was emphasized (De Weerd & Klandermans, 1999; Kelly & Breinlinger, 1996; Reicher 1984; Simon et al., 1998; Stryker, Owen & White, 2002). Lately, we see a growing interest in how emotions fuel protest participation (see Jasper, 1997; 1998; Goodwin, Jasper & Polletta, 2001; van Zomeren, Spears, Fischer & Leach, 2004). Goodwin et al. (2001, p. 1) were wondering how “academics have managed to ignore the swirl of passions all around them in political life”. In the first half of the last century, emotions were at the centre of protest studies. 2 This section bears heavily on Klandermans & Stekelenburg (forthcoming). CHAPTER 1 6 As a reaction to these irrational and emotional explanations, the dominating academic political analyses on protest participation then shifted to rationalistic, structural and organizational explanations; but, by reducing protest participation to a structural and rational process, researchers appear to have swung the pendulum too far in the opposite direction. As a result, emotions as explanations of protest were neglected altogether. Recently, it has been acknowledged that with the shift from irrational to rational the baby was thrown out with the bathwater. Indeed, the rational trend has now been reversed and we see emotions back on the research agenda of social movement scholars. We refer to the three social psychological approaches as instrumentality, identity and group-based anger. Each approach gives a different answer to the question of why people participate in protest campaigns, namely, people participate: (a) because they see it as an opportunity to change, at affordable costs, a state of affairs with which they are unhappy; (b) because they identify with the others involved; and (c) because they want to express their anger towards a target that has violated their values (Klandermans, 2004). Identity here receives more attention than the other approaches, because, as it turns out, identity plays a vital role in our model explaining protest participation. 1.2.1 Instrumentality In a reaction to the more traditional social psychological breakdown theories, resource mobilization theory stresses the importance of structural factors and underscores the rationality of participation in social movements. The unit of analysis of resource mobilization theory is essentially the meso or macro level. About 20 years ago, resource mobilization theory was the dominant approach in social movement research. This theory was blamed for being too structural and it went “too far in nearly abandoning the social-psychological level of analysis of social movements” (Klandermans, 1984, p. 584). Moreover, resource mobilization theory “underestimated the significance of grievances and ideology as determinants of participation in social movements” (Klandermans, 1984, p. 584). Additionally, in their decision to participate, people take reactions of others into account; indeed, costs and benefits are not evaluated in a social vacuum. To overcome these problems, Klandermans (1984) presented social-psychological expansions of resource mobilization theory that explain why some aggrieved people do and others do not participate in protest. His model reads as a plea for putting the social psychological level of analysis back into mobilization and participation theories. After all, participation in protest is individual behaviour. In this elaboration of the rational portrayal of mankind by INTRODUCTION AND THEORETICAL BACKGROUND 7 resource mobilization theory, the perceived costs and benefits of participation became the foundation of the model. Expectancy value theory and collective action theory are merged in the model. Expectancy value theory explains the motivation for specific behaviour by the value of the expected outcomes of that behaviour (Klandermans, 1984). The foundation of the socialpsychological expansion of resource mobilization theory is an individual’s expectation that specific outcomes will materialize multiplied by the value of those outcomes for the individual. In line with expectancy-value approaches (Feather & Newton, 1982), expectations and values stand in a multiplicative relationship. Protest participation is explained by the following motivational parameters: collective benefits and social and non-social selective incentives. Collective benefits are compounds of the value of the action goal and the expectations that the goal will be reached. These expectations in their turn are expectations about the behaviours of others, expectations that the action goal will be reached if many others participate and the expectation that one’s own participation will increase the likelihood of success. Expectancy value theory is based on a motivational analysis of a decision situation, possible outcomes, consequences of outcomes, and action taken. In collective action theory (Olson, 1965), then, people are conceived of as rational decision makers faced with a collective action dilemma: if the collective good is produced people will reap the benefits anyway, while the production of the collective good is not contingent on their own behaviour but on the joint efforts of the collective. Collective action theory predicts that under those circumstances rational actors will choose to take a free ride, unless selective incentives (i.e., those incentives that depend upon participation) motivate them to participate. As the decision to participate must be taken without knowing in advance the actual behaviour of others, individuals must rely on expectations about that behaviour (Klandermans, 1984). Indeed, since in the pre-protest period information about other people’s actual behaviour is often absent, potential protesters must rely on expectations about other people’s behaviour. Klandermans (1988) argues that free riding is less a problem than collective action theory suggests. He maintains that non-participation is only a free ride if the collective action is expected to be successful; but people are aware of the fact that some minimum number of participants is needed for an action to be successful. If people expect that this threshold will not be reached, they will refrain from participation, not because they are taking a free ride but because they expect that the collective action will fail. If, on the other hand, the threshold is reached, a few free riders more or less do not really matter. People’s views on whether the threshold will be passed are based on expectations about the behaviour of other people and, CHAPTER 1 8 according to Klandermans, such expectations operate as self-fulfilling prophecies: expecting a low attendance leads to decreasing motivation to participate, whereas high expectations increase motivation; thus expectations influence attendance. Klandermans (1984) distinguishes between three different motives for social movement participation is his model; each motive originates from different types of costs or benefits. The first motive, the collective motive, derives from the movement’s collective goals. In line with expectancy-value approaches (Feather, 1982), these blocks are conceptualized as the multiplicative function of the subjective value of the goal(s) of the movement and the subjective expectation that these goals will be reached. Expectations and value stand in a multiplicative relationship, because it is assumed that neither of the two terms can be nil. The second motive, the social motive, is derived from expected reactions of significant others to one’s participation in protest. It is conceptualized as the multiplicative function of the subjective (positive or negative) quality of others’ expected reactions and the personal importance of these reactions. Finally, the reward motive results from the selective incentives pertaining to more personal costs and benefits, such as paying for a train ticket or having a good time with friends. This motive is conceptualized too as a multiplicative function of value and expectancy components. The theory has found empirical support in several studies on a variety of movements (cf. Klandermans, 2003). Moreover, in a comparison of three movements (the labour movement, the women’s movement and the peace movement) Klandermans (1993) was able to show that movements differ in terms of what motives trigger participation. He argued that the orientation of the action for which each of the three movements was mobilizing—a strike, women’s groups in the community and a peace demonstration —appeals to different participation motives. He defined action orientation in terms of Turner and Killian’s (1987) description of action orientations that can determine the course of a mobilization campaign. Turner and Killian distinguish three action orientations: (1) power orientation, or an orientation toward acquiring and exerting influence; (2) participation orientation, or an orientation toward the benefits of participation; and (3) value orientation, or an orientation toward the goals and the ideology of the movement. Although, according to Turner and Killian, all three orientations play some role in every mobilization campaign, one orientation is always likely to dominate. Because strikes are power-oriented, they appeal specifically to the expectation component of the collective benefits. And, as Klandermans (1993) shows, the expectancy component was an important component in explaining trade unionists’ willingness to strike. In participation-oriented actions like the INTRODUCTION AND THEORETICAL BACKGROUND 9 women’s groups, women participated because participation in itself is perceived as satisfying. Indeed, most women referred to selective incentives as motives to participate in the women’s groups. The pattern of the peace movement’s demonstration yields the reverse pattern to the pattern of the strike. In this value-oriented demonstration, the expectancy component did not influence willingness to participate, but the value component carried great weight. Combining collective action theory with expectancy value models appeared to be a useful approach for the systematic analyses of the variety of beliefs, expectations and attitudes relating to participation in a social movement (Klandermans, 1984) and the comparison of motivational parameters dependent on action orientation (Klandermans, 1993). However, scholars doubted whether movement participation could be fully explained by rational considerations (Kelly & Breinlinger, 1996; Klandermans, 1986; 2003; 2004; Schrager, 1985). Kelly and Breinlinger (1996) argue that the assumption of rationality is especially strained in cases of protracted disputes. In these cases, union members often bear the financial and social burden of extreme hardship and are usually fully cognizant of the fact that benefits, if gained at all, may be slight. Again, a major limitation of this account is its neglect of the social and ideological aspects of collective action. As Schrager (1985, p. 859) points out, “collective action is more than the sum of economistic calculations: social and ideological factors figure powerfully in people’s willingness to act”. 1.2.2 Identity From various angles, the significance of collective identity as a motive stimulating participation in collective action was emphasized (Kelly & Breinlinger, 1996; Reicher, 1984; Simon et al., 1998; Stryker et al., 2002). Be it identification with workers or women (Kelly & Breinlinger, 1996; Kelly & Kelly, 1994), the elderly or gays (Simon et al., 1998, Simon & Stürmer, 2004), farmers (De Weerd & Klandermans, 1999), former East Germans (Mummendey, Kessler, Klink & Mielke, 1999) and the obese (Stürmer, Simon, Loewy & Jörger, 2003). All these studies have one basic assumption: the stronger the identification with a social category, the stronger the participation (or intention to participate) in collective action on behalf of that category. Identification with the group at stake seems a powerful reason to participate in protest on behalf of that group. What processes underlie identification and how do they relate to protest participation? To answer these questions we will first elaborate on the concept of identity. Thereafter, we will focus on a social psychological model of movement participation which distinguishes between an instrumental and an identity pathway to participation. CHAPTER 1 10 Identity. In the most basic social psychological sense, identity is a place in the social world (Simon, 1999). A place is a metaphorical expression and stands for any position on any socially relevant dimension such as nationality, ethnicity, gender, age and so forth (Simon, 1999). At the psychological heart of social identity theory (Tajfel & Turner, 1979) lies the assumption that people strive for a positive self-evaluation (Turner, 1999, p. 8). This self-evaluation encompasses two components: a personal and a social identity. Personal identity refers to selfdefinition in terms of personal attributes, whereas social identity refers to self-definition in terms of social category memberships. Social identity is seen as a cognitive mechanism; that is to say, if social identity becomes more salient than personal identity, then people see themselves less as unique individual persons and more as the prototypical representatives of their ingroup category. Indeed, people are inclined to define their personal self in terms of what makes them different from others, whereas they tend to define their social identities in terms of what makes them similar to others. In other words, it is the cognitive redefinition from an “I” into a “we” as a locus of self-definition that transforms individual into collective behaviour. When social identity becomes more salient than personal identity, people think, feel and act as members of their group (Turner, 1999). In the striving for a positive self-evaluation it is important that the membership of groups has a positive influence on one’s self-evaluation. Therefore people want to be members of high status groups. Because people strive for a positive self-evaluation they will, when confronted with a low group status, undertake action in order to acquire a high group status. Tajfel and Turner (1979) formulate the circumstances contributing to intergroup behaviour. Intergroup behaviour requires a low perceived group status and strong group identification (i.e., commitment to the group). In general, people do not feel committed to a low status group, but some social structural characteristics make commitment to a low status group viable. The first social structural characteristic indicated in social identity theory is permeability of the group boundaries, that is, the possibilities perceived by the individual to attain membership of a higher status group. When people conceive membership of a higher status group as a possibility, there will be no commitment to the lower status group. However, when people do not conceive possibilities to join a higher status group, people can feel commitment to the lower status group. The second social structural characteristic mentioned by social identity theory is stability. Stability is the extent to which status positions are stable or variable. People who conceive status positions as variable see collective action strategies as a possible method to realize higher group status. This perception implies that people are inclined to participate in collective actions on behalf of the group. This inclination will be enforced when the low group status is perceived as illegitimate or INTRODUCTION AND THEORETICAL BACKGROUND 11 unjustified. To sum up, according to social identity theory, people who feel strongly committed to their low status group will participate in collective action when they believe that collective action will change the low group status and when the low status is perceived as illegitimate. Group identification as the link between collective and social identity. Acting collectively requires some collective identity or consciousness (Klandermans & De Weerd, 2001). Collective identity and social identity are related concepts. However, they refer to different aspects of group life. Collective identity concerns cognitions shared by members of a single group, whereas social identity concerns cognitions of a single individual about his or her membership in one or more groups. Indeed, collective identity concerns shared beliefs and social identity concerns the incorporation of those socially shared beliefs (Klandermans & De Weerd, 2001). Klandermans (1997) made the distinction between the social construction of collective beliefs, which is the process of the formation of collective beliefs at the group level, and the appropriation of collective beliefs, which is the process of the formation of the idiosyncratic remakes of those beliefs at the individual level. Group identification, then, forms the link between collective and social identity. Group identification is a product of self-categorization—a cognitive representation of the self as an embodiment of a more inclusive category, accompanied by an awareness of similarity, ingroup identity and shared fate with others who belong to that category, the extent to which the ingroup is valued and self-involving (Brewer & Silver, 2001). Selfcategorization theory proposes that people are more prepared to employ a social category in their social identity the more they identify with that category. Thus, the stronger the group identification, the more the shared beliefs and fate comprised in the group’s collective identity are incorporated in the social identity. A group’s collective identity can be studied in its own right by examining such phenomena as the group’s symbols, rituals, beliefs and the values its members share. An individual’s identification with a group can be studied in its own right as well by examining the individual’s beliefs, sentiments, commitment to the group, use of symbols, participation in rituals and so on. Thus, group identification can be assessed in all kinds of ways, but any operationalization of group identification will refer somehow to what it means to an individual to belong to the group in point and will thus implicitly or explicitly refer to the pride of being a member of the group, to the symbols, the values, the fate shared by the group members. Therefore group identification is akin to commitment to the group (Klandermans & De Weerd, 2001, but see Kelly & Breinlinger, 1996; Mummendey et al, 1999). Huddy (2001), however, argues that it is not group identification per se but the strength of this identification that influences group members’ readiness to view themselves and act in terms of their group CHAPTER 1 12 membership. Huddy (2001) criticizes social identity literature for neglecting the fact that realworld identities vary in strength; but, she argues, identifying more or less strongly with a group may make a real difference, especially in political contexts. Salience. The fact that people have many collective identities raises the question of why some collective identities become central to mobilization while others do not, and why this is the case. People have many group memberships that remain latent most of the time. Selfcategorization theory (Turner, Hogg, Oakes, Reicher & Wetherell, 1987) hypothesizes that depending on contextual circumstances the transition from an “I” to a “we” as locus of selfdefinition occurs. A particular group membership and the associated identity is said to be salient to the extent to which it is “functioning psychologically to increase the influence of one’s membership in that group on perception and behavior” (Turner et al., 1987, p. 118). Social identities are social comparative and reality-based representations of self and others, and hence come into being and become meaningful in the social context and consequently will vary with the social context. Thus, salience is context dependent; it is a dynamic outcome of categorization in context (Turner, 1999). What turns a “sleeping identity” into a salient social identity and spurs action preparedness on behalf of that identity? Besides contextual factors and direct reminders, the presence of other ingroup members can be a potent reminder as well, even more so if the members are aiming at a common goal. Another effective prompt is being a minority. Although all of these reminders can make a social identity accessible, probably the most powerful factor that brings group membership to mind is conflict or rivalry between groups. The social identity approach suggests that salient social identity spurs several socialpsychological processes that facilitate group-serving behaviour. For example, when group members define themselves in terms of their collective identity, they focus on the similarities between themselves and fellow ingroup members with references to experiences, needs, interests or goals. As a result, “my” experiences and “your” experiences, needs and so forth are transformed into “our” experiences and needs. Group members’ perception that they share problems or grievances, or that their needs, goals and interests are interchangeable, is an important first step toward collective social and political action (Simon & Klandermans, 2001; Stürmer & Kampmeier, 2003). Politicized collective identity. Awareness of a collective identity does not necessarily make that identity politically relevant; therefore, collective identity must politicize (Simon & Klandermans, 2001). Politicization of collective identity and the underlying power struggle unfold as a sequence of politicizing events that gradually transform the group's relationship to its social environment. Typically, this process begins with the awareness of shared grievances. Next, INTRODUCTION AND THEORETICAL BACKGROUND 13 an external enemy is blamed for the group's predicament, and claims for compensation are levelled against this enemy. Unless appropriate compensation is granted, the power struggle continues. If in the course of this struggle the group seeks to win the support of third parties such as more powerful authorities (e.g., the national government) or the general public, collective identity fully politicizes (Simon & Klandermans, 2001). A collective identity of members of a particular group is politicized to the extent that “group members (self)-consciously engage in a power struggle of their group knowing that it is the wider, more inclusive societal context in which this struggle has to be fought out" (Simon & Klandermans, 2001, p. 319). A politicized collective identity is related to Tajfel’s (1971) concept of social change orientation (solving group problems through group actions), in that it indicates the process of investing the self in the group and can be understood as a form of collective identity that underlies group members’ explicit motivations to engage in such a power struggle. What distinguishes a politicized collective identity from a collective identity? The first distinction is consciousness raising: “the growing awareness of shared grievances and a clearer idea of who or what is responsible for those grievances reflect a distinct cognitive elaboration of one’s worldview providing group members with a meaningful perspective on the social world and their place [as group members, JvS] in it” (Simon & Klandermans, 2001, p. 327). The second distinction concerns the unique behavioural consequences of politicized collective identity, namely, politicized group members should be likely to engage in collective action directed at the government or the general public to force them to intervene or to take sides. All in all, identification with a group makes people more prepared to act as a member of that group, and therefore influences protest behaviour on behalf of that group. Simon and colleagues (Simon et al., 1998; Simon & Stürmer, 2004, Stürmer et al., 2003) acknowledge the influence of group identification on protest participation and develop a dual path model consisting of an instrumental and an identity pathway to protest participation. Dual path model to protest participation with an instrumental and identity path. Simon and colleagues (1998) propose a dual path model to protest participation: a calculation pathway, guided by instrumental reasoning that concentrates on the costs and benefits of participation, and an identification pathway that is guided by identity processes. The calculation pathway is represented by Klandermans’ (1984) instrumental model. Identity is elaborated in the context of social identity theory and is conceived in terms of Tajfel’s (1978, p. 63) definition of social identity: that is, the cognitive importance of the membership, the personal evaluation of the membership, and the emotional significance (Stürmer et al., 2003). CHAPTER 1 14 Stürmer et al. (2003) attempt to give a theoretical account of why strong identification with a group makes participation on behalf of that group more likely. According to Stürmer et al. (2003, p. 73), this is due to “a felt inner obligation to behave as a ‘good’ member”. These authors show that, when self-definition changes from personal to social identity, the group norm of participation becomes salient (Brown, 1980); the more one identifies, the more weight this group norm will carry and the more it will result in an “inner obligation” to participate on behalf of the group. In several studies exploring participation motives for various movements, Simon and his collaborates find empirical support for their concept of a dual pathway to social movement participation. Be it in their previously mentioned studies of identification with the Fat Acceptance Movement (Stürmer et al., 2003), the older people's movement or the gay movement (Simon et al., 1998), both calculation and identification made unique contributions to the prediction of willingness to participate. Rather than replacing instrumentality as an explanatory paradigm, identity added to the explanation as a second pathway. In fact, both calculation and identification could work together at the same time in the same person’s motivation. Identity processes appear to have both an indirect and a direct effect on protest participation (Stürmer, 2000, cited by Klandermans, 2000): a direct effect because collective identity creates a shortcut to participation: participation stems not so much from the outcomes associated with participation but from identification and solidarity with other group members involved (Klandermans, 2000); an indirect effect because collective identity influences instrumental reasoning, making it less attractive to take a free ride: high levels of group identification increase the costs of defection and the benefits of cooperation. 1.2.3 Group-based anger Emotions in protest research are new comers with a long history. Observing protesters makes one wonder why, until recently, protest researchers ignored emotions as explanations for protest behaviour. The sociologists Jasper (1997; 1998), Goodwin and Poletta (Goodwin, Jasper & Poletta, 2001) were the first to emphasize the importance of emotions in the context of political protest. They hold that emotions are important in the growth and unfolding of social movement and political protest. A central characteristic of emotions is social construction, in other words, the influence of norms, values and cultures on the experience of emotions. Goodwin et al. (2001) argue that emotions are socially constructed, but that “some emotions are more [socially, JvS] constructed than others, involving more cognitive processes” (p. 13). In their view, emotions that are politically relevant are, more than other emotions, at the social construction end of the scale. INTRODUCTION AND THEORETICAL BACKGROUND 15 For these emotions, cultural and historical factors play an important role in the interpretation of the state of affairs by which they are generated. The social construction of emotions is also an important aspect of the social psychological approach taken by van Zomeren et al. (2004) to collective action. Dual path model to protest participation with an instrumental and emotion path. Interestingly, the notion of a dual pathway was also proposed by van Zomeren et al. (2004) in their approach to collective action participation. These authors propose instrumentality and group-based anger as two pathways to protest participation. The theoretical foundation of van Zomeren et al.’s (2004) model is in the appraisal theory of emotion (Lazarus, 1991). The appraisal theory of emotion conceives appraisal, emotion and action as the means by which people perceive and cope with events in their social world. The theory distinguishes between problem-focused and emotion-focused coping. According to Lazarus (2001, p. 48), problem-focused coping is aimed at “obtaining information on which to act and mobilize actions for the purpose of changing reality”, whereas “the emotion-focused function is aimed at regulating the emotions tied to the situation”. Following Smith (1993), van Zomeren et al. (2004) propose an extrapolation of the appraisal theory of emotion to the group level, because group members who perceive disadvantages as collective appraise events in group rather than individual terms. By integrating various explanations offered in previous theory and research on protest participation, van Zomeren et al. (2004) created a dual path model, consisting of an instrumental and a group-based anger path, to explain protest participation. Group efficacy and action support play a central role in the instrumental pathway. Group efficacy is one’s collective belief that group-related problems can be solved by collective efforts (Bandura, 1997). Action support implies willingness of other group members to engage in collective action. Unfairness or illegitimacy and social opinion support play a central role in the group-based anger pathway. Following social psychological grievance literature, van Zomeren et al. hold that it is more often the procedures that lead to unfair outcomes than the outcomes per se that upset people. In addition to such procedural unfairness, social opinion support is proposed as the mechanism that helps to define the experienced unfairness as collective and shared. Social opinion support refers to the group members’ appraisal that their fellow group members share their opinion about the experienced unfairness. Appraisals such as unfairness (Miller, 2000) and emotional social support (Mackie, Devos & Smith, 2000) are believed to promote collective action because they evoke emotions such as anger (Smith, 1993). Van Zomeren et al. (2004) relate problem-focused CHAPTER 1 16 coping to the instrumental pathway to collective action and emotion-focused coping to the emotion pathway. Van Zomeren et al.’s (2004) model shows the importance of emotions as motivators in spurring action intentions, and once again the emotion pathway does not replace the instrumental pathway. Furthermore, the dual path model with an instrumental and emotion pathway to collective action implies that both pathways go together and in fact reinforce one another. If people can engage in both emotion-focused and problem-focused coping rather than only emotion-focused coping, their collective action tendencies tend to be higher. Indeed, perceiving instrumental support may also enhance perceptions of emotional support; however, perceiving emotional support does not necessarily enhance instrumental support (van Zomeren et al., 2004). Perceived action support and group efficacy enhance the expectation that other group members share your opinion, why would they otherwise (be willing to) participate in protest? But shared social opinion (i.e., emotional social support) does not necessarily reinforce perceived group efficacy (i.e., instrumental social support). Indeed, group efficacy is an intergroup perception (to what extent is my group able to change the intentions of another group?), whereas emotional social support is an intragroup perception (to what extent do group members share my opinion?). Therefore, it makes sense that the perception that other group members share one’s opinion (i.e., intragroup emotional support) does not necessarily make the group a powerful opponent in the political arena (i.e., an intergroup perception about power and status). The model reveals yet another reason why free riding is less a problem than suggested by Olson (1965). Olson’s protesters were just rational beings. Simon et al. (1998) show that identification processes help to overcome free riders’ dilemmas, while the emotional path points to emotion regulation as a mechanism to overcome free riding. After all, “the emotion-focused coping function is aimed at regulating the emotions tied to the situation” (Lazarus, 1991, p. 48), and one way to regulate these personally experienced emotions is to participate in collective action. Therefore emotional-focused coping makes free riding less likely, because one might take a free ride on the production of a collective good, but one cannot take a free ride on one’s own personal emotions. It is true that one way to regulate these personal experienced emotions is to participate in collective action, and expressing one’s view with like-minded people may function as an emotional valve, but if and only if one participates. A basic assumption underlying the model is that collective disadvantages lead to a salient identity leading to five group-based building blocks: social action support, group efficacy, procedural unfairness, social opinion support and group-based anger. Since the group plays such an important role in the model, one may wonder what the influence of identity strength might be. INTRODUCTION AND THEORETICAL BACKGROUND 17 Identification processes influence both instrumental reasoning (McCoy & Major, 2003) and emotions experienced as a group member (Yzerbyt, Dumont, Wigboldus & Gordijn, 2003). Therefore, it seems reasonable to expect that identification with the group involved influences both the instrumental and emotional pathway to protest participation. Moreover, as indicated, identifying more or less strongly makes a real difference, especially in political contexts (Huddy, 2001). In addition to identity salience, we propose to give identity strength a more prominent place in models predicting protest participation. We will return to this when we introduce our integrative model of the instrumental, identity, ideology and group-based anger paths. Now that we have briefly discussed the state of the art with regard to social psychological approaches to protest behaviour, we turn to our own approach to protest participation. 1.3 Introduction of a new social psychological approach to protest participation. Social psychologists, so far, have proposed three possible pathways that are relevant for participation in political protest: instrumentality, identity and group-based anger. Our own approach to protest participation is aimed at improving and expanding on these approaches. The findings with regard to both dual path models show that independent paths of, respectively, instrumentality and identity, and instrumentality and emotions, account for a significant proportion of the variance in willingness to participate in collective action. We propose that these approaches are not competing paradigms, but rather that the one compensates for the weaknesses of the other. This suggests that a model that encompasses the three pathways—instrumental, identity and emotional—might explain more variance in predicting protest behaviour than each motive separately or any possible twin. Indeed, one of the objectives of this dissertation is to integrate the three motives in a single model. Nonetheless, we are not completely satisfied with the three pathways alluded to in the literature. First of all, we feel that an important element is missing. In his discussion of fundamental motives for people to participate in social movements, Klandermans (2003; 2004) refers to ideology in addition to instrumentality and identity (in 2004 the group-based anger path appeared in the literature). Klandermans was neither the first nor the only one to mention these three fundamental motives explaining motivation. Indeed, “it has a long history in functional theories of attitudes and behaviour (see Sears, 1979; Sears & Funk, 1991)”, so states Klandermans (2003, p. 701). Recently, related triads have been proposed as antecedents of attitude importance (Boninger, Krosnick & Berent, 1995) and cooperative behaviour (Tyler & CHAPTER 1 18 Blader, 2000). Therefore we hold that, in addition to instrumentality, identity and group-based anger, ideology can explain a person’s motivation to participate in protest too. This would, then, make for four motivational pathways: people with an instrumental motive, who participate because they see it as an opportunity to change, at affordable costs, a state of affairs with which they are unhappy; people with an identity motive, who participate because they identify with the others involved; people with a group-based anger motive, who participate because it helps them to regulate the emotions tied to the situation; and people with an ideology motive, who participate because they want to express their view (Klandermans, 2004). In the next section, we elaborate on these pathways. If one considers various paths individuals might take on their way to participation, the question arises as to what factors determine which path is taken. This brings us to choices people make and decisions people take. The approaches discussed all make assumptions and theorize about human motivation. Therefore, an elaboration of the nature of motivation will be useful. When and why does someone take one pathway rather than another? Will a specific pathway prevail in some contexts? The answer to these kinds of questions presupposes a steering mechanism. Hence, in addition to our four possible routes to collective action, we conceive of a steering mechanism that helps to explain why one route to protest participation will prevail over another. Regulatory focus is the mechanism we propose. Regulatory focus refers to the process of self-regulation that can be pursued either with approach-oriented means or with avoidanceoriented means Higgins (1997; 1998). As previously mentioned, self-regulation dominated by strategic means that are approach-oriented is called promotion-focused, and self-regulation dominated by strategic means that are avoidance-oriented is called prevention-focused. Self-regulation can be seen as an adjustment of the self to a changing social and political environment. Indeed, participation because of common interests requires a shared interpretation of who should act, why, and how. Therefore, social movements do their utmost to explain how they interpret a situation. Social movement scholars conceptualize this process of meaning construction as framing; but, as there is an abundance of frames in the environment, why would one movement’s campaign be more appealing than another’s? We will investigate whether regulatory focus determines which collective action frame will be persuasive for whom. As mentioned in the introductory part of this chapter, a basic assumption holds that a match between a movement’s frame and the person’s regulatory state—a situation Higgins (1997; 1998) calls regulatory fit—will make a frame more appealing to that person. We will investigate whether organizations that frame a protest activity in promotion terms attract more promotion-focused than prevention-focused protesters, and whether organizations that frame a protest activity in INTRODUCTION AND THEORETICAL BACKGROUND 19 prevention terms attract more prevention-focused than promotion-focused protesters. In other words, is regulatory fit the mechanism that explains the chemistry between the individual protester and the mobilizing context? Next to regulatory fit as an explanatory principle for why a specific mobilization campaign appeals to one person rather than another, regulatory focus might be the reason why individuals tend to take different paths to protest. We will examine whether regulatory focus “controls” which path individuals take to protest. For example, are promotion- rather than prevention-focused protesters inclined to take the ideology pathway to collective action? In summary, we propose that a social psychological approach to political protest participation contains the following building blocks: participation motives and regulatory focus as a motivational steering mechanism making individuals inclined to take one route rather than another. In the following sections, we first elaborate on the four paths to protest participation and propose an integrative model resulting from existing theories on protest literature. Thereafter, we extend this model by including a motivational component. Therefore, we elaborate on regulatory focus theory. Finally, we integrate regulatory focus theory and our proposed protest participation model and hypothesize on which path will prevail in what social movement context, and how individual regulatory focus might control the pathway taken to political protest participation. 1.3.1 Instrumental path to protest participation. The instrumental path is theoretically rooted in Klandermans’s (1984) social psychological expansion of resource mobilization theory, namely, that protest participation with an instrumental goal stems from the assessment that some aims (social or political change) are attainable, at affordable costs, through collective action participation. In other words, protest participation is seen as a rational choice following from the expectation that protest will yield certain outcomes and the values of those outcomes. This participation motive resembles the instrumental pathway to collective action participation proposed by Klandermans (1984), Simon et al. (1998) and van Zomeren et al. (2004). As defending personal or group interests is the rationale to participate, purposefulness is central in this motive. When people take the instrumental path to political protest, they participate “for the purpose of changing reality” (Lazarus, 1991, p. 48). In other words, people taking the instrumental route are involved in problem-focused coping oriented toward instrumental strategies expected to improve their situation (see van Zomeren et al., 2004). Collective action is seen as an instrumental strategy to improve the situation of the group. CHAPTER 1 20 1.3.2 Identity path to protest participation. An identity motive refers to the circumstance that people identify with the others involved. This participation motive is theoretically rooted in the identity pathway to collective action proposed by Simon et al. (1998). People participate not so much because of the outcomes associated with participation but because they identify with the other participants. For people taking the identity path to protest participation, the focus changes from what “I” want to what “we” want (Brewer & Gardner, 1996). Collective action participation is seen as a way to show who “we” are and what “we” stand for, and people experience commitment and solidarity with other members of the group. Moreover, group members have the idea that “we” have much in common (by way of shared grievances, aims, values or goals). 1.3.3 Group-based anger path to protest participation. Raising one’s voice can be seen as emotional coping (van Zomeren et al., 2004). In other words, participating in collective action because of group-based anger is aimed at regulating the emotions tied to the social or political event. Thus, participating in collective action can be seen as an individual emotional catharsis, that is, a purging of emotions through expression, which makes participating in collective action because of a group-based anger motive a goal in itself. This path is theoretically rooted in the group-based anger path of van Zomeren et al. (2004). 1.3.4 Ideology path to protest participation. The ideology path to protest participation refers to people’s values and the assessment that these values have been violated. As Klandermans (2004, p. 365) states: “People are angry, develop feelings of moral indignation about some state of affairs or some government decision, and wish to make that known. They participate in a social movement not only to enforce political change but to gain dignity in their lives through struggle and moral expression”. A fundamental assumption on which this path relies is that people’s willingness to participate in political protest depends to a significant extent on their perception of a state of affairs as illegitimate (see van Zomeren et al., 2004), in the sense that it goes against fundamental values. An individual’s personal set of values is believed to strongly influence how, for example, a proposed policy, its ends and means, is perceived and evaluated. According to Rokeach (1973, p. 5), a “value is an enduring belief that a specific mode of conduct or end-state of existence is personally or socially preferable to an opposite or converse mode of conduct or end-state of existence. A value system is an enduring organization of beliefs INTRODUCTION AND THEORETICAL BACKGROUND 21 concerning preferable modes of conduct or end-states of existence along a continuum of relative importance”. For Schwartz (1992, p. 4), “values (1) are concepts of beliefs, (2) pertain to desirable end states or behaviors, (3) transcend specific situations, (4) guide selection or evaluations of behavior and events, and (5) are ordered by relative importance. Values, understood this way, differ from attitudes primarily in their generality or abstractness (feature 3) and in their hierarchical ordering by importance (feature 5)”. In principle, then, the distinction between attitudes and values is clear. “Attitudes refer to evaluations of specific objects while values are much more general standards used as basis for numerous specific evaluations across situations” (Feldman, 2003, p. 481). Hence, conceptualized in this manner, values are individual phenomena about which people usually feel strongly. They defend them in a variety of ways and react strongly when their values are challenged or frustrated (Feather & Newton, 1982). Indeed, “values are standards employed to tell us which beliefs, attitudes, values, and actions of others are worth challenging, protesting, and arguing about, or worth trying to influence or change” (Rokeach, 1973, p. 13). Therefore we assume that participating in collective action based on an ideology motive is grounded in a perceived violation of one’s values. As already stated, raising one’s voice can be seen as emotional coping (van Zomeren et al., 2004). Participating in collective action is—in addition to integrity maintenance—aimed at regulating the emotions tied to the social or political event. Thus, participating in collective action can also be seen as an individual emotional catharsis, which is a purging of emotions through expression. Both integrity maintenance and emotional catharsis make participating in collective action out of ideological goals a goal in itself. We argue that moral integrity maintenance and emotional coping work as an inner obligation. Maintaining one’s moral integrity and emotional coping each may incite an inner moral obligation to oneself, versus the inner social obligation to other group members incited by group identification. Indeed, besides identification and emotional coping, moral integrity maintenance might serve as a third mechanism to overcome the free riders’ dilemma built into the instrumental pathway. 1.3.5 An integrative model of the instrumental, identity, ideology and group-based anger paths. In social movement research, little is known about the relative weight of instrumentality, identity, ideology and group-based anger (Klandermans, 2003). From the research of Klandermans (1984), Simon and colleagues (Simon et al., 1998; Stürmer & Simon, 2004, Stürmer et al., 2003) and van Zomeren et al. (2004), we know that at least instrumentality, identity and group-based anger CHAPTER 1 22 paths have their impact, but how do they work together? Do they add to each other, or do they interact? Are they correlated or independent determinants of participation? The findings with regard to the dual path models of Simon and colleagues and van Zomeren et al. show that independent paths of, respectively, instrumentality and identity, and instrumentality and groupbased anger, account for a significant proportion of the variance in the willingness to participate in collective action. In the protest literature so far, three independent pathways to protest participation come to the fore, but we propose ideology as a fourth path. We are not aware of studies investigating all three pathways (let alone all four pathways) to protest participation in one study, but the results of the two dual path models suggest that instrumentality, identity and group-based anger represent three independent pathways to protest participation, and we presume that ideology functions as an independent path as well. One may, of course, also wonder whether the four are in fact independent. We presume that this is unlikely, if only because the group-based emotions literature available in social psychology suggests otherwise. Take, for example, intergroup emotions theory (Smith, 1993; 1999; Mackie et al., 2000) from where the group-based anger path of van Zomeren and colleagues (2004) theoretically stems. In an attempt to move beyond an individualized context, Smith (1993; 1999) builds upon self-categorization theory (Turner et al., 1987) and proposes a model of social emotions in which people are thought to be able to experience emotions on behalf of their group. According to intergroup emotions theory, people not only categorize themselves as group members, but care about situations or events affecting the ingroup as well. Since intergroup emotions theory is based on the presumption that the group is incorporated in the self (“the group is in me”, thus “I feel for us”), one would assume that the more the group is in me (i.e., the higher the identification with the group), the more people experience group-based emotions. Smith and colleagues have not applied this litmus test, since they use the salience paradigm. Yzerbyt et al. (2003), however, did, and showed that emotional reactions fully mediated the impact of categorization context and identification on action tendencies. In other words, the salience of similarity was found to generate angry feelings among participants only to the extent that they strongly identified with the relevant category. Thus, people will experience group-based emotions when the social category is salient and when they identify with the group at stake. The findings of Yzerbyt et al. (2003) might point to a protest participation model that departs from a model with four independent paths. After all, these findings suggest that groupbased emotions mediate between group identification and action tendencies, which, translated to our model, might suggest that group-based anger mediates on identity motives and the INTRODUCTION AND THEORETICAL BACKGROUND 23 willingness to participate in collective action. Indeed, the dual path model of van Zomeren and colleagues (2004) allows for the influence of social identity on protest participation through the concept of identity salience. However, the findings of Yzerbyt and colleagues (2003) suggest that, in a model based on identity strength, the group-based anger path might mediate between identity motives and motivation to participate. Nonetheless, based on the evidence of Simon and colleagues (Simon et al., 1998; Stürmer & Simon, 2004; Stürmer et al., 2003) we assume that both identity and instrumental motives influence motivational strength directly as well. Moreover, as indicated, we hypothesize that ideology motives will have a direct effect on motivation too. Therefore, the first step in building a model that will account for protest participation is to propose a model explaining protest participation comprised of four direct pathways: instrumental, ideology identity, and group-based anger and one indirect pathway: identity via group-based anger. The next step in building our model is based on the assumption that group identification influences the reasons why people participate in political protest. Identity motives as moderator. Group identification is an awareness of similarity, ingroup identity and shared fate with others who belong to the same category (Brewer & Silver, 2001). It has pervasive effects on what people feel, think and do (Terry & Hogg, 1996). According to self-categorization theory (Turner et al., 1987), people are likely to self-stereotype as a function of their level of identification with the group. Strong identifiers are expected to adopt the prototypical behaviours of the ingroup to a greater extent than weak identifiers. In fact, when identification with the group is strong, group interests are internalized and adopted as self-interests (Turner et al., 1987). Tyler and Bladel (2003, p. 352) maintain precisely the same as social categorization theory: “identity judgments will be the primary factors shaping attitudes, values, and cooperative behaviours in groups”. If identity judgments shape attitudes, values and cooperative behaviours (Turner et al., 1987; Tyler & Bladel, 2003), they might shape the reasons why people participate, in other words, their participation motives. In statistical terms, we hypothesize that identity motives moderate the relation between instrumental and ideology motives. Instrumental and ideology motives as mediator. Group members are not just angry, they are angry about some state of affairs or some government decision. We presume that groups can be angry for instrumental and ideology reasons. Therefore, we hypothesize that identity motives influence group-based anger motives indirectly via their CHAPTER 1 24 influence on instrumental or ideology motives. In statistical terms, we hypothesize that instrumental and ideology participation motives mediate the relation between identity and groupbased anger motives. Integrated mediator-moderator model. Figure 1.1 displays the model as developed in the previous section. The model aims to account for motivational strength, that is, the strength of the motivation to participate. Although many people may sympathize with the goals of social movements often only a small percentage actually participate in collective action to achieve these goals. Klandermans & Oegema (1987) developed a model of social movement participation addressing this issue. According to this model becoming a participant in a social movement can be conceived of as a process involving four different steps: First, people have to sympathize with a movement and thus become part of the mobilization potential; second, they have to be targeted by mobilization attempts; third, they must develop the motivation to participate in particular collective actions, and forth and finally they must overcome possible barriers to participate. The in this thesis presented research examines the motivational strength of people actually participating and not the often applied motivation to participate in future protest events. In other words, the present research focuses primarily on Step 3 and 4 of the Klandermans’ model. Sampling those people that actually participate, implicates, by definition, that these people are motivated and overcame possible barriers. However, this does not necessarily imply that all people ending up participating were equally motivated. On the contrary, one may assume that the strength of the motivation of the people who end up participating varies. It is this variation that we are interested in. instrumentality group-based anger identity motivational strength ideology Figure 1.1. Integrated mediator-moderator model accounting for motivational strength to participate in political protest. INTRODUCTION AND THEORETICAL BACKGROUND 25 In the first place, the model depicts the four direct paths from instrumental, identity, ideology, and group-based anger motives to motivational strength. In addition to these direct paths, several indirect paths are mapped. The mediator model assumes that identity motives influence groupbased anger motives indirectly via their influence on instrumental or ideology motives. The moderator model assumes that weak and strong identifiers differ in their inclination to employ instrumental or ideological motives. As a consequence, instrumental and ideological participation motives are stronger predictors of group-based anger for strong rather than weak identifiers. Indeed, identity plays a major role in the model. The most important prediction of this model is that for strong identifiers, group identification influences group-based anger motives via its influence on instrumental or ideology motives. For weak identifiers, however, identity motives will be of minor importance, therefore we hypothesize that weak identifiers will start at the instrumental or ideological pathway directly. In other words, identity motives moderate the mediators (instrumental and ideological motives). Social movement context as moderator. Earlier we distinguished three action orientations (power, value and participation) that can determine the course of a mobilization campaign, as outlined by Turner and Killian (1987). Following Klandermans (1993), we assume that, depending on their action orientation, movements appeal to different participation motives. “The more power-oriented a campaign is, the more strongly it will emphasize the movement’s effectiveness, its ability to exert influence. Therefore, a movement must convince the individual that the planned action will be successful. If it does not succeed in convincing potential participants—if its effectiveness seems dubious —an individual will have little reason to participate” (Klandermans, 1993, p. 389). Because participation on the basis of an instrumental participation motive implies that participation is seen as an opportunity to change a state of affairs at affordable costs, we assume that a power-oriented protest event will be appealing to people with an instrumental motive. “The more value-oriented a campaign is, the more it will emphasize the importance of its goals and the ideology behind them” (Klandermans, 1993, p. 389) and the more it will give participants the opportunity to express their discontent with a given state of affairs. Because participation on the basis of an ideology motive is aimed at expressing one’s views and venting one’s anger against a target that has violated one’s values, we assume that protest events with a value-action orientation will be appealing to people with ideology motives. “Participation-oriented campaigns emphasize the opportunity for individuals to engage in activities that are satisfying in and of themselves. Accordingly, a participation-oriented campaign CHAPTER 1 26 appeals to selective incentives” (Klandermans, 1993, p. 389). In participation-oriented actions, people participate because participation in itself is perceived as satisfying; therefore we presume a direct route from identity motives to motivational strength. Translated into the mediator-moderator model presented here, we hypothesize that, in the context of power-oriented protest, the group-based anger of high identifiers is reinforced by instrumental motives, identity motives moderate instrumental motives, which in their turn mediate between identity and group-based anger. In the context of value-oriented protest, the group-based anger of high identifiers is reinforced by ideology motives, identity motives moderate ideology motives, which in their turn mediate between identity and group-based anger. And, although we have no data on participation-oriented protest events, we assume that identity motives will influence motivational strength directly in the context of participation-oriented protest. After all, participation with others involved in itself is perceived as satisfying, suggesting that participants need no other motivational force to be spurred into participation. In summary, following Klandermans (1993), we assume that, depending on their action orientation, movements appeal to different participation motives. As a result, our mediatormoderator model, in its turn, will be moderated by the social movement context. As a consequence, we hypothesize that group identification (i.e., identity motives) moderate different mediators in power- and value-oriented protest. In building our model we started off with four possible routes to protest participation: instrumental, identity, ideology and group-based anger motives. Our next step was to theorize about the possible effects of identity salience versus identity strength on group-based anger. Based on findings of Yzerbyt et al. (2003), we concluded that, in a model accounting for protest participation, the effects of identity strength (i.e., identity motives) might be mediated by groupbased anger. This resulted in four direct paths to protest participation: instrumental, identity, ideology, and group-based anger and one indirect: identity via group-based anger. The next step in our reasoning suggested that weak and strong identifiers differ in their inclination to employ instrumental or ideology motives. This introduced the moderating effects of identity on instrumental and ideology motives. Subsequently, we introduced a mediation hypothesis, holding that identity motives influence group-based anger motives indirectly via their influence on instrumental or ideology motives. Then, we combined the moderation and mediation hypothesis into an integrated mediator-moderator model accounting for the strength of motivation to participate in political protest. Finally, we introduced the moderating effects of the social movement context on the mediator-moderator model: the social movement context may influence which path will be taken. INTRODUCTION AND THEORETICAL BACKGROUND 27 We will test the proposed model in the context of two different social movement campaigns: a power-oriented and a value-oriented campaign. As a consequence we will not be able to test our hypotheses regarding participation-oriented campaigns. The proposed model, so far, operates within an existing theoretical framework derived from protest literature. The last step in building our theory is to account for regulatory processes controlling the paths taken to protest participation. We presume that regulatory focus influences how people interpret and evaluate their (political) environment and explains why one route to protest participation will prevail over another. Indeed, we presume that regulatory focus moderates the paths taken to protest participation. 1.4 Regulatory Focus Theory What makes people take one possible action rather than another? Indeed, what makes people participate in one possible collective action rather than another? These questions refer to motivation. If one were to choose a single term for what is meant by motivation, it would be the verb “to want”. This one term captures in everyday language a wide variety of relevant meanings, for example: to have or feel need of, to be necessary (require); to wish or demand the presence of; to desire to come, go, or be; to have a strong desire for or inclination to (like); to fail to possess (lack). What motivates people to do or not to do something in the context of what is important to them is in (social) psychological terms a matter of self-regulation. Self-regulation refers to the process by which people formulate objectives to attain personal goals, plans, or standards and apply strategies to realize those goals (Mischel, Cantor & Feldman, 1996). As previously mentioned, Higgins proposed a motivation theory based on this self-regulation principle: regulatory focus theory (Higgins, 1997; 1998). Given that the root of the term “motivation” is “to move”, it is not surprising that a classic way of thinking about motivation is in terms of approach (moving toward) and avoidance (moving away from). This so called hedonic principle underlies motivational models across all levels in psychology: for example, the appetitive system involving approach and the aversive system involving avoidance (Lang, 1995) and approach tendencies of individuals with a “hope of success” and the avoidance tendencies of individuals with a “fear of failure” (Elliot & Trash, 2002; 2004). It is clear from these models and the empirical support they have acquired that people are motivated to approach pleasure and to avoid pain. In regulatory focus theory, approach and avoidance is conceptualized in terms of strategic means for self- CHAPTER 1 28 regulation. Higgins argues that self-regulation can be pursued through means that are either approach-oriented or avoidance-oriented. However, according to regulatory focus theory, people employ substantially different strategies when they approach pleasure and when they avoid pain. Self-regulation dominated by strategic means that are approach-oriented is called promotionfocused, and self-regulation dominated by strategic means that are avoidance-oriented is called prevention-focused. These two modes of self-regulation are the source of different goals, needs and values and have distinctive cognitive, emotional and behavioural consequences for how people respond to their social environment (Higgins, 1998). Figure 1.2 gives a schematic representation of Regulatory Focus Theory. - avoidance strategic mean - ought goals - security needs - tradition/conformity values - no-loss / loss - vigilance - approach strategic mean - ideal goals - nurturance needs - self-direction/stimulation values - gain/no-gain - eagerness prevention-focus promotion-focus actual state Figure 1.2. A schematic representation of regulatory focus theory (Higgins, 1997; 1998) with the goals, needs and values triggering, respectively, a prevention-focus and a promotion-focus. Individuals can differ in their chronic promotion-focus on hopes, aspirations and accomplishments versus chronic prevention-focus on duties, obligations and safety. Differences in chronic regulatory focus can arise from differences in the quality of parental involvement (see Higgins & Silberman, 1998). A child-parent relationship characterized by encouraging accomplishments and withdrawing love as discipline produces strong ideals, representing hopes and aspirations and promotion concerns with accomplishments and advancements. In contrast, a history of protection and using punishment as discipline produces strong oughts, representing duties and obligations and prevention concerns with safety and security (see Higgins & Silberman, 1998). In addition to varying chronically across individuals, regulatory focus can vary INTRODUCTION AND THEORETICAL BACKGROUND 29 across situations. Regulatory focus can be induced temporarily in momentary situations. Just as the responses of caretakers provide promotion or prevention feedback, task feedback or task instructions concerning which actions will produce which consequences can induce regulatory focus by communicating gain/non-gain information (promotion) or non-loss/loss information (prevention). More generally, gain and non-gain situations will induce a promotion-focus, whereas non-loss and loss situations will induce a prevention-focus. We propose that self-regulation can be seen as an adjustment of the self to a changing social and political environment. Recently, studies relating self-regulation to intergroup behaviour (Levine, Higgins, & Choi, 2000; Sassenberg, Kessler & Mummendey, 2003; Shah, Brazy & Higgins, 2004) have shown that self-regulation processes apply to both personal and social identity. Therefore, we hold that social and political changes may be perceived as a threat to one’s personal or group-based goals, needs and values. Given the far reaching and diverse implications of the two self-regulatory foci on perceiving the social and political world, we assume that they also influence the reason(s) why people are motivated to participate in political protest. We hold that motivation to participate in protest for prevention-focused and promotionfocused people stems from threats to fundamentally different goals, needs and values. Therefore we presume that regulatory focus influences who participates in collective action, and when and why they do so. Because regulatory focus theory has been extensively covered elsewhere (e.g., Higgins, 1997; 1998), we will only discuss the distinct goals, needs and values relating to the two foci and regulatory fit. 1.4.1 Distinct goals, needs and values. A common denominator of goals, needs and values is that they motivate people. The regulatory foci control which goals, needs and values motivate whom. Goals. Motivation stems from attaining personal or group goals. Self-regulating behaviour is the process by which people initiate, adjust, or stop actions in the pursuit of such goals. Promotion-focused self-regulation is more likely in the pursuit of goals that are related to advancement and accomplishment (ideal goals). Promotion-focused people are concerned with the presence or absence of positive outcomes and their strategic inclination is approach in a state of eagerness. Successfully attaining an ideal goal is perceived as a gain, whereas failing to attain an ideal goal is seen as a non-gain. Prevention-focused self-regulation, on the other hand, is more likely in the pursuit of goals that are related to security and protection (ought goals). Preventionfocused people are concerned with the presence or absence of negative outcomes and their CHAPTER 1 30 strategic inclination is avoidance in a state of vigilance. Successfully attaining an ought goal is perceived as a non-loss, whereas failing to attain an ought goal is perceived as a loss. Needs. A felt need is a “force that organizes perception, apperception, intellection, conation and action” (Ronen, 1994, p. 242). These forces create a state of tension that an individual attempts to relieve through appropriate actions. The theory of self-regulatory focus begins by assuming that the hedonic principle should operate differently when serving fundamentally different needs, such as the distinct survival needs of nurturance (nourishment and psychological growth and development) and security (e.g., protection and safety). Nurturancerelated regulation involves a promotion-focus, whereas security-related regulation involves a prevention-focus. Values. Violated values are strong motivators (Feather & Newton, 1982; Rokeach, 1973; Schwartz, 1992). Promotion-focused people are motivated by a violation of self-direction values (independent thought and action choosing, creating and exploring) and stimulation values (excitement, novelty and challenge in life, Van-Dijk & Kluger, 2004). Prevention-focused people, in contrast, are motivated by violation of tradition values (respect, commitment, and acceptance of the customs and ideas that traditional culture or religion provide), conformity values (restraint of actions, inclinations, and impulses likely to upset or harm others and violate social expectations or norms) and security values (safety, harmony and stability of society, of relationships and of self, Van-Dijk & Kluger, 2004). 1.4.2 Regulatory focus as moderator of the paths taken to protest participation. Given the fundamentally different goals, needs and values that motivate prevention-focused and promotion-focused people, we hypothesize that regulatory focus moderates the reasons why people are motivated to take part in collective action. In other words, does regulatory focus moderate the paths taken to protest participation? How may these distinct goals, needs and values relate to the proposed participation motives? Because prevention- and promotion-focused people are concerned with pursuing specific goals and fulfilling distinct needs, we propose that promotion- and prevention-focused protesters with an instrumental participation motive want to solve distinct problems by protest participation. This leads to the first assumption: prevention-focused people are motivated to take the instrumental pathway when security needs are threatened, and protection and responsibility goals are obstructed; and promotion-focused people are motivated to take the instrumental pathway when nurturance needs are threatened, and growth and advancement goals are obstructed (hypothesis 1). INTRODUCTION AND THEORETICAL BACKGROUND 31 Assumptions regarding an identity pathway follow from the distinction between ought and ideal goals. As described, strong identification with a group makes participation on behalf of that group more likely due to a felt inner obligation to behave as a “good” member (Stürmer et al., 2003). Because oughts (which tend to trigger a prevention focus) involve duties, obligations and responsibilities that generally are interpersonal, whereas ideals (which trigger a promotionfocus) involve aspirations that are often personal (Pham & Higgins, 2005), we propose that a group norm to participate in collective action (i.e., employ an identity participation motive) will carry more weight for prevention-focused people than for promotion-focused people. In other words, we hypothesize that prevention-focused people are more inclined to take the identity pathway to protest participation than promotion-focused people (hypothesis 2). Prevention- and promotion-focused people prioritize also differential values. Therefore we assume that the motivation to take the ideology path to political protest stems from violation of distinct values: prevention-focused people are motivated to take the ideology pathway to collective action when tradition, conformity or security values are violated, and promotionfocused people when self-directed and stimulation values are violated (hypothesis 3). Prevention- and promotion-focused protesters are both angry, although for different reasons. Therefore, we assume that for prevention-focused protesters and promotion-focused protesters group-based anger motives are equally important (hypothesis 4). 1.4.3 Regulatory fit: social movement context and regulatory focus. Regulatory focus theory predicts that information that is consistent with the regulatory focus should have more impact than information that is inconsistent with this focus (Higgins, 1997; 1998). Two types of mechanisms have been proposed to account for this match between information and the person’s regulatory state—a situation Higgins calls regulatory fit. First, it could be that a promotion- or prevention-focus calls attention to information that is compatible with this state and increases the weight that this information receives during judgment integration. If the weight of regulatory-focus-compatible information increases in judgment, options that are attractive on compatible dimensions will naturally be evaluated more favourably. Consistent with this explanation, Aaker and Lee (2001) found that, following exposure to a promotional message, respondents had a better memory for information that was consistent with their state of regulatory focus than for information that was inconsistent with this state. Pham and Avnet (2004) arrived at the conclusion that information compatible with the regulatory state was perceived to be more diagnostic than information that was not compatible. CHAPTER 1 32 An alternative explanation is that regulatory fit creates a subjective experience of “feeling right” that is then used as information to make evaluations (Higgins, 2000; Higgins, Idson, Freitas, Spiegel, & Molden, 2003). Cesario, Grant and Higgins (2004) recently tested this explanation in a persuasion context. They hypothesized and found that, for a promotion-focused person, a message framed in eager terms feels more “right” than a message framed in vigilant terms. Conversely, for a prevention-focused person, a message framed in vigilant terms feels more “right” than a message framed in eager terms. These feelings of rightness are then interpreted as meaning that the message is persuasive or that the person agrees with the message’s position. They additionally found that this effect disappeared when the actual source of the feelings was made salient before message exposure. This latter finding supports the idea that the phenomenon is driven by a misattribution of the feelings of rightness to the object being evaluated (see Schwarz & Clore, 1983, cited by Pham & Higgins, 2005). The regulatory fit hypothesis may influence the persuasiveness of a collective action frame. Although our social and political environment provides an abundance of frames, regulatory fit may explain when and why people adopt certain frames while neglecting others. We hold that social or political change framed in promotion terms will attract more promotionfocused people, whereas social or political change framed in prevention terms will attract more prevention-focused people (hypothesis 5). Studies on regulatory fit have found various effects of regulatory fit. Regulatory fit increases the intensity or strength of activity engagement, which in turn can increase performance. When people imagine prospective events during the decision-making process, regulatory fit increases the imagined pleasure of positive events and the imagined pain of negative events. When a decision is made with regulatory fit, people are more satisfied with the decision and evaluate it more favourably. They also “feel right” about their decision and believe that what they did was right and fair. Finally, the value experienced on making a decision with regulatory fit is transferred to the outcome of the decision, such that the outcome is perceived as more valuable. These regulatory fit effects have been shown to be independent of just the positive or negative mood of the decision makers (Higgins, 2000; Higgins et al., 2003). 1.5 This study In conclusion, collective action is a collective response to a collectively experienced threat. We assume that individual self-regulation proceeds from a reaction to threatened but valued ought or ideal group goals, needs and values. Subsequently, these self-regulation processes are assumed to INTRODUCTION AND THEORETICAL BACKGROUND 33 influence what collective action frame will appeal to whom. At the individual level, it is assumed that self-regulation processes control which path protesters take to participation in political protest events. Thus, as a consequence of chronic self-regulatory processes, people are inclined to take the instrumental, identity or ideology path to protest participation. Moreover, prevention-focused protesters and promotion-focused protesters are both angry, although for different reasons. Testing the hypotheses based on the argumentation so far, will be dealt with in the following chapters. 34 Chapter 2. Development and validation of RFQ-proverb. Two types of measures of regulatory focus can be distinguished3. The first type is based on the assumption that chronic ideal goals or ought goals, that is self-guides, involve a promotion-focus or prevention-focus, respectively. The presumption is that the stronger the ideal or ought goals, the stronger the corresponding regulatory focus. Accordingly, these measures indirectly assess regulatory focus. In this first type of measure, Selves Questionnaires are applied (Higgins, 1987) and a measure is derived from these questionnaires, the Strength-of-Guide measure. The second type of measure uses questionnaires and is developed to tap into regulatory focus directly. Two Regulatory Focus Questionnaires (RFQ) have been published. Higgins et al. (2001) designed the first RFQ to assess individuals' subjective histories of success in promotion and prevention self-regulation strategies. The second was designed to measure promotion and prevention goals in relation to positive or negative role models (Lockwood, Jordan & Kunda, 2002). However, in our search for a reliable and valid RFQ we ran into several problems (which we will elaborate on in the next section) with the existing measures; this made us decide to develop a new instrument. In this chapter we describe the development and validation of this instrument. First we give an overview of the existing assessment techniques to measure chronic regulatory focus, that is, the Selves Questionnaires, Strength-of-Guide measure and the two RFQs. This overview is confined to chronic regulatory focus because this is what we wish to measure. Next, we introduce the new measure—RFQ-proverb—and elaborate on the rationale for using proverbs to measure regulatory focus. The remainder of the chapter concerns three studies designed to scale characteristics, reliability, construct validity and predictive value of RFQproverb. 2.1 Available assessment techniques to measure chronic regulatory focus. 2.1.1 Selves Questionnaires The use of Selves Questionnaires is grounded in the foundation of regulatory focus theory, selfdiscrepancy theory (Higgins, 1987). Self-discrepancy theory has been concerned with self-guides as chronic goals, and actual-self matches and mismatches to self-guides as chronic goal attainments, that is, chronic successes and failures, respectively (see Higgins, 1989; 1996a). The theory postulates that, when the individuals’ represented actual selves are congruent with their self-guides, they feel good. When they are discrepant, they feel bad. Self-discrepancy theory USING PROVERBS TO MEASURE REGULATORY FOCUS 35 supposes two cognitive dimensions that underlie various self-state representations. These dimensions are domains of the self and standpoints of the self. According to self-discrepancy theory there are three basic domains of the self: (a) the actual self, that is, one’s representation of the attributes that one believes one actually possesses; (b) the ideal self, one’s representation of the attributes that one ideally would like to possess (i.e., a representation of one’s hopes, wishes or aspiration); and (c) the ought self, which is one’s representation of the attributes that one should or ought to possess (i.e., the representation of one’s sense of duty, obligations or responsibilities). Paring each standpoint on the self with a domain of the self creates the six basic kinds of self-state representations addressed by self-discrepancy theory: actual-own, actual-other, idealown, ideal-other, ought-own and ought-other. The actual self-representations (both own and other) constitute what is typically meant by a person’s self-concept, and the remaining self-state representations are self-directive standards or acquired guides for being. Self-discrepancy theory proposes that people are motivated to reach a condition in which their actual state matches their ideal or ought states; that is, where their selfconcept matches their self-guides (Hook & Higgins, 1988). Selves Questionnaires (Higgins, Bond, Klein, & Strauman, 1986) is a measure of selfdiscrepancy that measures the extent to which current actual states match ideal and ought selfguides. Respondents are asked to list eight or ten attributes for each of the three different selfstates: (a) their actual self, (b) their ideal self and (c) their ought self. The questionnaire is typically administered in two sections, the first involving the respondent’s own standpoint and the second involving the standpoint of the respondent’s significant others (i.e., mother and father). The procedure for calculating the magnitude of an ideal discrepancy or ought discrepancy involves comparing the actual self attributes to the attributes listed in either their ideal self or their ought self to determine which attributes in the actual self match or mismatch the attributes of that particular self. The self-discrepancy score is basically the number of mismatches minus the number of matches. The ideal self-discrepancy score represents the extent to which an individual is currently failing in promotion self-regulation, whereas the ought self-discrepancy score represents the extent to which an individual is currently failing in prevention self-regulation (Higgins et al., 2001). Selves Questionnaires is a method that, though widely used, has been criticized for being too difficult for participants; this may limit its ability to obtain unique self-representations. It is also difficult and time-consuming to score (Tangney, Niedenthal, Covert & Barlow, 1998); this 3 This chapter draws on Stekelenburg & Klandermans (2003). CHAPTER 2 36 makes such a method less suitable for research like ours in a field setting. Furthermore, in Selves Questionnaires participants think of their different selves in terms of descriptive attributes and do not reflect directly on discrepancies (or congruencies), let alone on the strategic means for regulating these discrepancies (i.e., regulatory focus). Because Selves Questionnaires are an indirect measure of regulatory focus, Strauman (1996) made a plea for measures that were more direct. An instrument measuring regulatory focus based on the strength rather than descriptions of the different selves is the Strength-of-Guide measure (see Higgins et al., 2003; Higgins, Shah & Friedman, 1997; Shah & Higgins, 1997). 2.1.2 Strength-of-Guide measure Like the Selves Questionnaires, the Strength-of-Guide measure asks participants to list attributes of the “selves”, but the latter is a computer measure that focuses on the response time of every given attribute instead of discrepancies between actual and ideal/ought selves. The assumption underlying this method is that the response time required to produce a given attribute reflects how readily it comes to mind and, thus, its accessibility (Fazio 1986; 1995; Higgins et al., 1997). The operationalization assumes that (a) accessibility represents activation potential and (b) stored knowledge with higher activation potential produces faster responses. In other words, a selfguide with a high activation potential—thus often activated—produces faster responses than selfguides with a low activation potential. Furthermore, the Strength-of Guide-measure only focuses on ideal and ought self-representations. Thus, unlike the Selves Questionnaires, the standpoint of significant others and the above-mentioned actual selves are no longer taken into account. Respondents are asked to provide attributes describing their ideal and ought selves. Higgins et al. (1997) found a negative relation between the discrepancy and accessibility of the self-guide measured by the Strength-of-Guide measure, so that (a) an increasing accessibility of ideal self-guides reflects an increasing actual-ideal discrepancy and (b) increasing ought self-guides reflects increasing actual-ought discrepancy. Therefore Higgins et al. suggest operationalizing regulatory focus with the Strength-of-Guide measure. The Strength-of-Guide measure is still the standard to assess regulatory focus in the laboratory. Unfortunately, this operationalization lends itself only to research conducted with the aid of computers. Moreover, researchers and respondents see it as a “time-consuming reactiontime measure of regulatory focus” (Lockwood et al., 2002, p. 861). Both computerization and the time it consumes make this measure less suitable for research in a field setting. However, both Higgins et al. (2001) and Lockwood et al. (2002) developed easy to administer scales as a useful alternative to the computer assessed Strength-of-Guide measures. USING PROVERBS TO MEASURE REGULATORY FOCUS 37 2.1.3 Regulatory Focus Questionnaire by Higgins et al. In a recent series of studies, Higgins et al. (2001) developed and employed a measure of individuals’ subjective histories of effective and ineffective promotion and prevention selfregulation: the Regulatory Focus Questionnaire (RFQ). They discovered that participants with a subjective history of success in promotion or prevention self-regulation experienced a sense of achievement pride. Moreover, depending on the type of achievement pride, goal-oriented behaviour differed. Promotion achievement pride appeared to orient participants toward eagerness means to attain the goal, whereas prevention achievement pride oriented participants toward vigilance means. The promotion subscale measures individuals’ subjective histories of promotion success with items such as “How often have you accomplished things that got you ‘psyched’ to work even harder?” and “I have found very few hobbies or activities in my life that capture my interest or motivate me to put effort into them” (reverse scored). The prevention subscale measures individuals’ subjective histories of prevention success with items such as “How often did you obey rules and regulations that were established by your parents?” and “Not being careful has gotten me into trouble at times” (reverse scored). Higher scores on either the promotion or the prevention subscale reflect individuals’ sense of their history of promotion or prevention success in goal attainment, respectively. Higgins et al. (2001) report excellent scale characteristics and, subsequently, in several studies regulatory focus has been measured with this RFQ (see Ayduk, May, Downey & Higgins, 2003; Camacho, Higgins & Luger, 2003; Grant & Higgins, 2003). 2.1.4 Regulatory Focus Questionnaire by Lockwood et al. Nearly simultaneous with Higgins et al. (2001), Lockwood et al. (2002) created a second Regulatory Focus Questionnaire. In this RFQ, participants indicate the extent to which they endorse promotion goals (e.g., “I frequently imagine how I will achieve my hopes and aspirations”; “I often think about the person I would ideally like to be in the future”) and prevention goals (e.g., “I frequently think about how I can prevent failures in my life”; “I am anxious that I will fall short of my responsibilities and obligations”). Like the Higgins et al. (2001) RFQ, Lockwood et al. (2002) report excellent scale characteristics, and later studies where this scale was applied have been reported (see Lockwood, Marshall & Sadler, 2005; Lockwood, Sadler, Fyman & Tuck, 2004). Lockwood et al. (2002) do not explain their motives for creating a new RFQ. However, they do elaborate on the difference between their RFQ and Higgins et al.’s (2001) RFQ. “Higgins and his colleagues have also measured regulatory focus by examining individuals’ subjective CHAPTER 2 38 experiences of success in obtaining past prevention and promotion goals. Our measure of promotion and prevention was designed to tap into the theoretical underpinnings of promotion and prevention concerns directly, providing a concise means of assessing them” (Lockwood et al., 2002, p. 859). Thus, whereas Higgins et al. (2001) measure subjective experiences of successfully obtaining prevention and promotion goals, the Lockwood RFQ “assesses chronic promotion and prevention goals directly” (Lockwood et al., 2002, p. 859). 2.1.5 Possible limitations of the RFQs of Higgins et al. and Lockwood et al. Although the Higgins et al. (2001) and Lockwood et al. (2002) RFQs seem to be an improvement in terms of applicability compared to the Strength-of-Guide measure, we can still identify several aspects of the existing questionnaires that make them problematic to employ in our study. The most important problem relates to the population to which it has been administered. Both RFQs were administered to a student population, whereas we want to employ an RFQ in a field setting with average citizens. An item, for example, like “How often did you obey rules and regulations that were established by your parents?” may be applicable to students, but a union member of 78 years of age may be at least surprised to be asked such a question. The other problems we have with the available RFQs relate to the content of the separate items of the questionnaires. One of the problems is generalizability; if specific behaviour is taken as an indicator, the measure may not generalize well beyond the behaviours included in the questionnaire. Finally, Higgins et al. (2001) assert that the RFQ is carefully balanced in parental and non-parental content between promotion and prevention. In other words, the items refer to the socialization period with the parents. We doubt whether this is actually the case. Four of the five items of prevention refer to this socialization period, whereas none of the six items of promotion refers to it. Another set of problems concerns the need to translate instruments in non-English speaking countries. Indeed, the translation of instruments from one language to another is fraught with difficulties (Perkins et al., 2004). “Substantive problems can occur due to an unexamined transfer of concepts from one culture/language system to another and/or lack of equivalence in words used to express concepts in the two languages due to differences in affect, familiarity, and clarity. These can introduce serious biases into research, compromising the scientific integrity of the results” (Erkut et al., 1999, p. 207). We discern that direct translation of questionnaires is one of the most frequently used methods in social psychology in the development of non-English versions of instruments. However, direct translations have been “repudiated as an unreliable method for achieving language equivalence” (Erkut et al., 1999, p. 208). Ideally, the aim of a USING PROVERBS TO MEASURE REGULATORY FOCUS 39 translation is to achieve versions of an English instrument that are conceptually, rather than literally, equivalent. Therefore, the focus should be a concept-driven rather than a translationdriven approach to attain equivalence in instruments (Erkut et al., 1999). Considering these problems, we felt compelled to develop an RFQ that did not have these problems. In our search for a solution, it occurred to us that an RFQ consisting of proverbs could be the answer. Many proverbs/sayings4 are either promotion-oriented (e.g., “Risk your neck”) or prevention-oriented (e.g., “Better safe than sorry”). This observation became our guiding principle in developing an RFQ. In the following section we first elaborate on the rationale for an RFQ consisting of proverbs. Subsequently, we report on three studies to test the reliability, validity and predictive value of an RFQ-proverb we developed. 2.2 Proverbs as elements of an RFQ. Wolfgang Mieder (1995), a leading scholar of proverbs, defines proverbs as a concise statement of an apparent truth, which has currency among the people. Proverbs possess a number of characteristics that make them particularly suitable to measure regulatory focus. They reflect the assumptions, attitudes and motivations of the members of a culture precisely because they are absorbed at an early age and then are taken for granted (Wederspahn, 2002). Indeed, proverbs bear an interesting resemblance to regulatory focus. Both are absorbed at an early age, in other words during socialization. However, we may well expect that not all individuals absorb the same proverbs. In fact, we hold that a social regulatory style will be passed on byamong other thingsusing those proverbs that convey the norms and values related to the social regulatory style of the socializers. People appear to be especially sensitive to information that fits their dominant regulatory focus (Cesario et al., 2004; Higgins, 2000; 2002). This on its own justifies the assumption that someone’s dominant regulatory focus can be assessed with the help of proverbs. Some psychologists have explicitly rejected proverbs as candidates for ideas about human behaviour and their use in mental testing, mostly because proverbs too often represent contradictory ways of construing problem situations (McKeachie & Doyle, 1970). In Chapter 1 we saw the contradictory ways prevention-focused and promotion-focused people construe social situations; this is precisely what makes proverbs ideally suited for an RFQ. Take again, for example, the two contradictory proverbs “Better safe than sorry” and “Risk your neck”. The first immediately makes one think of prevention-focus and the second of promotion-focus. Indeed, CHAPTER 2 40 Pham and Higgins (2005) demonstrated that prevention-focused people use vigilance strategies, whereas promotion-focused people use eagerness strategies. As these examples suggest, proverbs lend themselves very well to be categorized as either prevention- or promotion-oriented. This is not surprising, because both regulatory focus and proverbs are handed down by human experience. Hence, if regulatory focus is such a general principle that it influences all spheres of life, these foci will inevitably be reflected in proverbs. Therefore, some proverbs should be more appealing to a person than others, depending on his or her chronic regulatory focus. Socialization, sensitivity to information and precipitation form the theoretical foundation for our rationale to use proverbs in the construction of an RFQ. At face value and on theoretical grounds, we expect that chronic regulatory focus can be assessed by an RFQ-proverb. It is not enough for an idea to be appealing however; evidence supporting the idea is needed as well. Therefore, we designed three studies to test the reliability, construct validity and predictive value of an RFQ based on proverbs. First, we elaborate on the scale development followed by the overall scale characteristics of the three studies. Then, we say more about the relation between RFQ-proverb and a translated RFQ (Lockwood et al., 2002) and the relation to theoretically related measures. Thereafter we describe the three studies separately. Study 1 was designed to test validity in a context that is known to differentiate between promotion regulatory focus and prevention regulatory focus. Study 2 was designed as a replication. Study 3 was designed to test whether accessibility of a dominant chronic regulatory focus could be measured by RFQ-proverb; this was done by relating score and response time to each other (Fazio, 1995). 2.3 Development and evaluation of RFQ-proverb. 2.3.1 Scale development We started the process of scale development by examining a Dutch search engine for proverbs containing 1,631 proverbs and sayings. After we had eliminated proverbs based on redundancy or because they did not express meaning relating to regulatory focus, 301 of the 1,631 remained. The 301 proverbs were checked for face validity. This gave us a set of 102 proverbs (57 relating to prevention and 45 relating to promotion). This set of 102 proverbs was given to two content experts. To ensure high content validity, that is, the extent to which the instrument evokes a range of responses relating to regulatory focus, the content experts were asked to categorize the proverbs according to the following classification: 4 Although RFQ-proverb is comprised of proverbs and sayings, in the text “proverbs” is used as shorthand for both. USING PROVERBS TO MEASURE REGULATORY FOCUS Promotion 41 Prevention • nurturance needs • security needs • growth and advancement goals • protection and responsibility goals • self-direction and stimulation values • tradition and conformity values • eagerness • vigilance • openness to change • need for stability • success = gain, failure = no-gain • success = no-loss, failure = loss • approach • avoidance The categorization of the content experts eventually gave us a set of 27 promotion proverbs and 30 prevention proverbs. Examples of promotion categorizations are: “Where there’s life, there’s hope” (ideal goals) and “Risk your neck” (risk seeking) and prevention categorizations: “Better safe than sorry” (protection and security) and “Let the cobbler stick to his last” (need for stability). The context experts’ categorization of the proverbs according to the different implications enabled us to create an instrument that tapped into a broad range of the rich implications of the different foci. Finally, out of these remaining proverbs, the ones most frequently used in day-to-day language were chosen. This resulted in an RFQ-proverb consisting of 15 promotion and 15 prevention proverbs (see Table 2.1). Table 2.1 Overview of prevention and promotion items of RFQ-proverb. promotion items prevention items Give it your all* Where there’s a will, there’s a way* You never know what you can do until you try* Life is for living* Variety is the spice of life* Broaden your horizons* Nothing ventured, nothing gained* Risk your neck Don’t hide your light under a bushel A cat in gloves catches no mice Our fate is in our own hands If you’re not in, you can’t win Where there’s life, there’s hope Never trouble trouble till trouble troubles you Not moving forwards means moving backwards Act normal, that’s crazy enough* Let the cobbler stick to his last* Prevention is better than cure* Don’t skate on thin ice* Sufficient unto the day is the evil thereof* Wait to see which way the wind blows* East west, home is best* Discretion is the better part of valour A burnt child dreads the fire A bird in the hand is worth two in the bush One swallow does not make a summer Keep something aside for a rainy day Better safe than sorry Haste makes waste Seeing is believing * the starred items are included in the short version of RFQ-proverb 42 CHAPTER 2 This RFQ-proverb was submitted in three separate studies to students from the Faculty of Social Sciences at the Vrije Universiteit, Amsterdam. In the first two studies, respondents filled out a paper-pencil version during a lecture break. In the third study, the data were obtained using the Internet as the means of data collection. We will elaborate on the methodology used in the three studies in the following section. Here it suffices to mention the instruction provided to the respondents. In all three studies, the instructions to the respondents were exactly the same. RFQproverb was introduced by stating that proverbs reflect the ways that people appraise the world around them. After this short introduction, respondents were instructed to answer the following question “To what extent do the following proverbs apply to all your actions?” on a 7-point Likert scale, ranging from 1 (not at all) to 7 (very much). The instructions were followed by an example, whereupon the respondents could indicate the extent to which the 30 proverbs applied to all their actions. In the following pages we elaborate on the scale characteristics of RFQ-proverb. Next, we examine the convergent validity by relating RFQ-proverb to a translated RFQ (Lockwood et al., 2002) and theoretically related measures. Finally, we describe the three studies separately. 2.3.2 Scale characteristics To examine the construct validity of RFQ-proverb, that is, the extent to which the instrument measures two constructs, we report characteristics of the distributions and factor analyses. Characteristics of the distributions The three studies produced very similar results. Table 2.2 shows the scale characteristics for promotion and prevention proverbs that emerged for the three samples. Table 2.2 Overall scale characteristics for promotion and prevention items over three studies. Study 1 (N = 109) promotion prevention Study 2 (N= 53) Study 3(N = 186) promotion prevention promotion prevention Mean 5.17 4.20 5.40 4.07 5.17 4.24 Median 5.17 4.25 5.47 4.13 5.20 4.27 Mode 4.83 4.17 5.67 3.80 5.27 3.93 SD 0.61 0.77 0.66 0.77 0.58 0.66 Skewness .14 (.23) -.63 (.18) -.18 (.18) -.38 (.23) -.63 (.33) -.07 (.33) USING PROVERBS TO MEASURE REGULATORY FOCUS Kurtosis -0.39 (.47) -0.33 (.47) 0.88 (.67) -0.59 (.67) 43 1.35 (.36) -0.54 (.36) Shapiro-Wilk’s .99 (ns) .98 (ns) .97 (ns) .99 (ns) .97 (.02) .99 (ns) Cronbach's alpha .78 .79 .83 .82 .78 .77 The mean, median and mode are almost identical, for prevention items very close to the midpoint of the scale (4) and for promotion items reasonably close. The standard deviations are very similar and close to 1. Kurtosis and skewness tests of normality were used to evaluate whether the RFQ-proverb data follow the theoretical normal distribution. The three studies show a rather varying pattern: in the first study prevention items are negatively skewed, whereas in the second and third study the promotion items are negatively skewed. Studies 1 and 2 show good to moderate kurtosis; however, in Study 3, promotion is too tall and prevention is too flat. These distribution normality findings indicate that the pattern is changeable; therefore we employed another goodness-of-fit test. Besides skew and kurtosis, another goodness-of-fit method to test normality of distribution is the Shapiro-Wilk's test. According to the Shapiro-Wilk’s test, all the subscales in all the studies have a normal distribution except the promotion subscale in Study 3. Given these results, we assume that the subscales provide for a close to normal distribution shape. In addition to satisfying distribution criteria, the internal coherence of the subscales appear to be high (Cronbach's alphas between .77 and .83). Moreover, male and female respond similarly to the scale items. Overall, then, the scale characteristics of RFQ-proverb appear to be good. Factor analyses Subsequently, we conducted Principal Component Analyses (PCA) with oblimin rotation allowing for natural correlation among the factors (following Higgins et al., 2001) to determine the extent to which promotion-focus and prevention-focus were represented by the proverbs in the RFQ-proverb. Table 2.3 gives an overview of the first six components resulting from each PCA, the natural correlation and percentage of declared variance. 44 CHAPTER 2 Table 2.3 Results from three PCAs with oblimin rotation of RFQ-proverb from Studies 1, 2, and 3. Eigenvalues Comp. natural correlation explained variance Study 1 Study 2 Study 3 RFQ short (N= 109) (N = 53) (N = 186) (N = 442) 1 4.51 5.45 4.80 3.18 2 4.05 4.94 3.64 2.63 3 2.28 2.13 1.88 1.19 4 1.77 1. 89 1.75 .99 5 1.57 1. 68 1.53 .97 6 1.45 1. 54 1. 34 .80 .03 .03 .05 .06 29 % 35 % 28 % 41 % On the basis of the scree test we concluded that each factor analysis revealed two factors with eigenvalues between 3.65 and 5.45, accounting for respectively 29%, 35%, 28% and 41%of the variance. The natural correlation between the two constructs is low (.03 - .06), indicating that the two vary independently (see Higginset al., 1986). We will return to RFQ-short later. Thus, prevention and promotion measured by RFQ-proverb can be treated as two separate constructs. After the PCAs with oblimin rotation, we conducted four orthogonal PCAs with varimax rotation to facilitate the interpretation. Based on the scree test, we restricted each PCA to two factors. Table 2.4 provides an overview of the factor loadings for the two factors on the 15 prevention and 15 promotion items. USING PROVERBS TO MEASURE REGULATORY FOCUS 45 Table 2.4 Factor loadings for 2 factors resulting from PCA on 15 prevention and promotion items (7 in RFQ short) in Studies 1, 2, and 3 and for RFQ-short. Prevention items Act normal that's crazy enough Let the cobbler stick to his last Prevention is better than cure Don’t skate on thin ice Sufficient unto the day is the evil thereof Wait to see which way the wind blows East west home is best Better safe than sorry Keep something aside for a rainy day Discretion is the better part of valour A burnt child dreads the fire A bird in the hand is worth two in the bush Haste makes waste One swallow does not make a summer Seeing is believing Study 1 Study 2 Study 3 RFQ-p short (N = 109) (N = 53) (N = 186) (N = 442) Comp Com Comp Comp Comp Comp Comp Comp p2 1 2 1 1 2 1 2 -.02 -.01 .07 -.01 -.03 .19 -.23 .01 .02 .00 .03 .13 -.01 .11 -.43 .58 .66 .20 .50 .09 .33 .56 .63 .60 .51 .61 .37 .46 .62 .40 -.01 .13 .04 -.03 .01 -.11 -.10 .15 .31 -.53 -.04 -.03 -.03 .02 -.31 .59 .42 .58 .77 .51 .59 .35 .68 .48 .26 .75 .70 .57 .58 .14 -.07 -.02 .12 .02 -.14 -.17 -.07 .11 .08 -.13 .10 .10 .11 .02 -.33 .55 -.16 .68 .51 .04 .70 .35 .23 .51 .59 .18 .65 .21 .09 .49 .43 .05 .61 .65 .05 .68 .43 .54 .56 .64 .43 .54 .42 .29 .30 .57 .73 .65 .60 .53 You never know what you can do until you try .37 Where there’s life, there’s hope .31 Our fate is in our own hands .41 Risk your neck .53 Don’t hide your light under a bushel .53 A cat in gloves catches no mice .67 If you’re not in, you can’t win .70 Never trouble trouble till trouble troubles .41 you Not moving forwards means moving .35 backwards -.18 .12 .01 .10 -.10 -.23 -.07 .24 .07 .11 .33 .05 -.12 .08 .45 .57 .66 .51 .45 .63 .43 .39 .55 .61 .73 .71 .84 .23 -.02 -.04 .02 .19 .07 -.27 .39 .13 .03 .07 .04 -.02 -.08 .02 .42 .68 .71 .68 .59 .67 .38 .45 .43 .44 .62 .26 .60 .27 .00 -.07 .00 .13 .24 -.15 -.06 .29 .14 -.13 .09 .22 -.03 .15 .05 .45 .07 .35 .08 Promotion items Variety is the spice of life Broaden your horizons Nothing ventured, nothing gained Where there’s a will, there’s a way Give it your all Life is for living 66 .69 .73 .60 .47 .73 .58 .02 .00 .02 .33 .18 -.05 .10 In each of the four studies the factor loadings of the 30 proverbs (or 14 in the short version) differentiated as expected over the two factors. These two factors are referred to as promotion and prevention. CHAPTER 2 46 Summing up, RFQ-proverb appears to have good distributions properties, while the results of the factor analyses consistently revealed the same factor structure over the three samples. 2.3.3 A short version We realized that, because of practical constraints, it could be difficult to administer the full version of RFQ-proverb in many settings. Consequently, we sought to develop a short version that would correlate highly with the total scale but would consist of substantially fewer items. To develop a short version of RFQ-proverb we selected the 14 (7 promotion and 7 prevention items) proverbs that had the highest item-total correlation. The proverbs included in RFQ-proverb-short are the starred proverbs in Table 2.1. RFQ-proverb-short has scale characteristics nearly identical to those of the complete scale5. See Table 2.3 for the eigenvalues and Table 2.4 for the factor loadings of RFQ-proverb-short. The scree test of the short version reveals two factors with eigenvalues between 3.17 and 2.63, accounting for, respectively, 41% of the variance. The natural correlation between the two constructs is low (.06), indicating that the two factors of RFQ-proverb-short vary independently. 2.3.4 Relationships with other scales To assess its relationship to comparable scales and instruments, that is, to assess the convergent validity, we administered some of these measures in Study 2 and Study 3 in addition to RFQproverb. To test the convergent validity, first of all we included a translated version of the RFQ created by Lockwood et al. (2002). Obviously, we anticipated a very strong relationship between RFQ-proverb and Lockwood et al.’s RFQ. Additionally, scales were included that, theoretically, were expected to correlate with the regulatory focus pair and thus with the subscales of RFQproverb. Personal Need for Structure (PNS, Neuberg & Newsom, 1993), Preference for Consistency (PFC, Cialdine, 1995), Need For Cognition (NFC, Cacioppo, Petty & Kao, 1984) and the Schwartz Value Scale (SVS, Schwartz 1992) were included. The Personal Need for Structure scale was created in an attempt to capture, as a chronic individual motive, several aspects of the desire for simple structure. 5 Scale characteristics of the promotion subscale of the RFQ-proverb-short are: M = 5.15, Median = 5.14, SD = 0.88, Cronbach's alpha = .76, skew = -.53 (ses = .12), kurtosis = 1.03 (sek = .24) and Shapiro-Wilk = .99 (.01). Scale characteristics of the prevention subscale of the RFQ-proverb-short are: M = 4.21, Median = 4.28, SD = 1.05, Cronbach's alpha = .73, skew = -.21 (ses = .12), kurtosis = .10 (sek = .24) and Shapiro-Wilk = .98 (.05). And, as with the total scale, male and female participants responded similarly to the RFQ-proverb-short. USING PROVERBS TO MEASURE REGULATORY FOCUS 47 People high on personal need for structure would prefer structure and clarity in most situations, with ambiguity and grey areas proving troublesome and annoying. Such people should experience discomfort if they perceive structure and clarity to be missing from situations. We assume promotion-focused people with their openness to change to be low on need for structure, and prevention-focused people with their preference for stability to be high on need for structure. The Preference for Consistency scale measures variation in the desire to be and to be seen as consistent. “Individuals who score low on the PFC Scale appear to prefer spontaneity, change, and unpredictability in their responding rather than congruency with their prior responses. They seem open and oriented to the new, in ways that are relatively unconstrained by the established. For those who score high on the scale, however, personal consistency is valued, and these individuals take pains to align their responses in most situations with their previous actions, attitudes, and commitments, especially when the concept of consistency has been made salient” (Guadagno Asher, Demaine, & Cialdini, 2001, p. 859). Following the same reasoning as the Personal Need for Structure scale we assume that promotion-focused people will be low on PFC and prevention-focused people high. The Need for Cognition scale assesses the extent to which people differ in their tendency to engage in and enjoy effortful cognitive activities (Cacioppo et al., 1984). Need for cognition is typically measured with the 18-item Need for Cognition scale (Cacioppo et al., 1984). Research using the scale has demonstrated that individuals with a high need for cognition prefer complex, cognitively demanding tasks to simpler versions; individuals with a low need for cognition prefer simpler tasks to more complex versions (Cacioppo et al., 1984). We administered a short version of the Dutch translation by Pieters, Verplanken & Modde (1987). Molden and Higgins (2004) show that, when faced with uncertainty, prevention-focused people have a low need for cognition, and promotion-focused high. We expect to find this relation as well. Finally, we were interested in the relation between RFQ-proverb and values. Van-Dijk and Kluger (2004) showed that promotion-focused people endorse self-direction and stimulation, whereas prevention-focused people endorse tradition, conformity, and security. Therefore we related RFQ-proverb to these values from the Schwartz Value Scale (SVS). Thus, if RFQ-proverb measures a prevention and promotion construct one would expect that promotion measured with RFQ-proverb would relate to self-direction and stimulation and not to tradition, conformity, and security. By contrast, prevention should relate to tradition, conformity, and security and not to self-direction and stimulation. 48 CHAPTER 2 It is usually recommended to control for the “alternative type of score” in regulatory focus research, because significant interrelations are typically found in studies based on regulatory focus (Higgins et al., 1997, p. 518). Although we found low insignificant correlations of the RFQ-proverb prevention and promotion subscales (.03 - .06), we followed up the recommendation of Higgins et al. (1997). Therefore, all the studies reported in this dissertation look at the unique relation between the specific regulatory focus and the outcome variable. In other words, the unique relation between promotion-focus and an outcome variable is determined, controlling for prevention regulatory focus, whereas the unique relation between prevention-focus and an outcome variable is determined, controlling for promotion-focus. Table 2.5 (Partial) correlations of the RFQ-proverb with other scales. Correlation Scale promotion prevention Partial Correlation promotiona preventionb N RFQ Lockwood promotion .76** .02 .76** .01 53 RFQ Lockwood prevention .04 .84** .06 .84** 53 .46** -.36** .52** 186 .33** .07 .32** 186 Personal Need for Structure -.22** Preference for Consistency .09 Need for Cognition .20** -.01 .20** -.02 186 SVS Self-direction .37** -.11* .39** -.16** 442 Stimulation .47** -.10* .49** -.17** 442 Tradition Conformity Security Note. a .05 .32** .02 .32** 442 .12* .33** .10* .33** 442 .09 .36** .09 .35** 442 = controlled for prevention, b = controlled for promotion *p <.05. **p<.01. RFQ-proverb proved to be very strongly related to RFQ Lockwood (see Table 2.5), thus the convergent validity is high. RFQ-proverb prevention is positively correlated to need for structure and preference for consistency, and RFQ-proverb promotion is positively correlated to need for cognition. Furthermore, RFQ-proverb prevention appears to be positively correlated to tradition, conformity and security, whereas RFQ-proverb promotion is positively correlated to selfdirection and stimulation. This is what one would expect if RFQ-proverb is to measure regulatory USING PROVERBS TO MEASURE REGULATORY FOCUS 49 focus. However, contrary to our expectations, RFQ-proverb promotion is positively correlated to conformity also, although RFQ-proverb prevention was much more strongly related to conformity than RFQ-proverb promotion. Thus, our expectations were confirmed for the most part. In summary, the results of the scale characteristics and the relationships with related measures are promising. Based on the observed relations with related concepts and RFQ Lockwood, we suggest that differences in promotion-focus strength and prevention-focus strength can be measured with RFQ-proverb in a reliable and valid way. Indeed, the scale differentiates between promotion-focused and prevention-focused people. The relation between RFQ-proverb and other scales is, however, no more than an important beginning in our search for validation. Above all, an instrument should have predictive validity. Therefore we made an effort to establish the construct validity through a series of studies. These studies were designed to examine the extent to which RFQ-proverb enabled us to differentiate between a chronic promotion and prevention regulatory focus and the extent to which these regulatory foci related differences in intrinsic and extrinsic study motivation, in other words the predictive value. 2.4 Study 1 and Study 2 As an additional test of the validity of RFQ-proverb, we tested in Study 1 and Study 2 whether RFQ-proverb differentiates between intrinsic and extrinsic study motivation. Students with an intrinsic motivation study because they find it inherently interesting or enjoyable, whereas students with an extrinsic motivation study because they experience control or pressure from others, for example their parents (see, for example, Deci & Ryan, 1991). Intrinsic and extrinsic motivation to study are related to different kinds of basic needs or incentives, such as distinguishing between (intrinsic) wants related to autonomy or self-determination and (extrinsic) wants related to control or pressure from others (Deci & Ryan, 1991). This fits with regulatory focus theory: prevention-focused students who aim at ought goals with an avoidance strategy should show a stronger extrinsic study motivation than promotion-focused students. Promotionfocused students who aim at ideal goals and adhere to values of self-direction and autonomy should show stronger intrinsic study motivation than prevention-focused students. Hence, oughts can be internal and ideals can be external as well; however, on average, we assume that oughts (which trigger a prevention-focus) evoke an extrinsic motivation, whereas ideals (which trigger a promotion-focus) evoke an intrinsic motivation. RFQ-proverb should reveal this relation. Thus a 50 CHAPTER 2 high score on RFQ-proverb promotion should relate to high intrinsic motivation to study, whereas a high score on RFQ-proverb prevention should relate to high extrinsic motivation to study. Study 2 was designed to replicate the findings of Study 1; therefore Study 1 and Study 2 are identical, except for the test of the convergent validity RFQ Lockwood, which was only conducted in Study 2. 2.4.1 Study 1 and Study 2 Methodology. Participants: Students (Study 1: 109 [28 male, 81 female], Study 2: 53 [15 male, 38 female]) from the Faculty of Social Sciences at the Vrije Universiteit, Amsterdam, participated in this study. During a lecture they were asked to participate (voluntarily and unpaid) in a paperpencil test framed as an achievement study motivation study. We found no significant gender differences. Materials: RFQ-proverb was used to assess chronic regulatory focus. The items to measure intrinsic and extrinsic study motivation were based on the situational motivation scale (SIMS, Guay, Vallerand & Blanchard, 2000). Both the intrinsic and extrinsic motivation scales were comprised of three items. Respondents were instructed to indicate on a 7-point Likert scale, labelled 1 (not at all) to 7 (very much), the extent to which the following study motivation items applied. The items included were: (1)“I study because I think studying is interesting”, (2) “I study because I think studying is pleasant” and (3) “I study because studying is fun” (intrinsic study motivation, Study 1: = .71, Study 2 = .87). (1) “I study because I feel that I have to do it”, (2) “I study because I am supposed to do it”, and (3) “I study because it is something that I have to do” (extrinsic study motivation, Study 1: = .67, Study 2 = .58). Procedure: The study was presented as an investigation of study motivation. During a lecture break, the class filled in a paper-pencil test, which took about ten minutes. Respondents completed RFQ-proverb and subsequently the SIMS. In addition to RFQ-proverb and the SIMS, the questionnaire included some demographic variable questions, such as age and gender. 2.4.2 Results and Discussion Study 1 and Study 2. Preliminary analyses Table 2.6 provides an overview of the preliminary analyses. The intercorrelations between study motivation and regulatory focus are in the expected directions. RFQ-proverb promotion is positively related to intrinsic study motivation and not to extrinsic study motivation, and RFQ- USING PROVERBS TO MEASURE REGULATORY FOCUS 51 proverb prevention is positively related to extrinsic study motivation and not to intrinsic study motivation. Table 2.6 Correlations, means and standard deviations for intrinsic study motivation, extrinsic study motivation, promotion-focus and prevention-focus for Study 1 and Study 2. Correlations of Study 1 are above the diagonal and those of Study 2 under the diagonal. intrinsic extrinsic promotion prevention -.20* .16* .08 .31** intrinsic -- extrinsic -.45** -- .01 promotion .21* -.07 -- -.05 prevention -.18 .24* .01 -- M Study 1 5.60 2.93 5.17 4.20 M Study 2 6.04 2.33 5.40 4.07 SD Study 1 0.67 1.35 0.61 0.77 SD Study 2 0.81 1.21 0.66 0.77 Note. * significant at .05 level (two-tailed), ** significant at .01 level (two-tailed) Main analyses For Study 1 and Study 2 we performed four hierarchical regressions. As indicated, it is usually recommended to control for the “alternative type of score” in regulatory focus research, because significant interrelations are typically found in studies based on regulatory focus. Moreover, the correlation between types of study motivation is significant (Study 1: r = -.20 and Study 2: r = .45); therefore, we controlled for the “alternative type of score” (Higgins et al., 1997, p. 518). Accordingly, each type of study motivation was regressed on the other type of study motivation, RFQ-proverb promotion and RFQ-proverb prevention. The first analysis involved the prediction of intrinsic study motivation from RFQ-proverb prevention and RFQ-proverb promotion, controlled for extrinsic study motivation. The second analysis involved the prediction of extrinsic study motivation from RFQ-proverb prevention and RFQ-proverb promotion, controlled for intrinsic study motivation. Subsequently, identical analyses were performed for Study 2. Table 2.7 gives an overview of the findings of the four analyses. 52 CHAPTER 2 Table 2.7 Results of four hierarchical regression analyses with intrinsic or extrinsic study motivation as outcome variable and RFQ-proverb/promotion and prevention as predictors controlled for either intrinsic or extrinsic study motivation. Study 1 (N = 109) intrinsic R2 Study 2 (N = 53) extrinsic R2 F intrinsic R2 F extrinsic R2 F F Model 1: intr./extr.a .03 2.92 -.17* .03 2.92 -.17* .21 .11 5.80***-.22* .03 12.70***-.45*** .20 12.70*** -.45*** Model 2: intr./extr.a .06 3.12* -.23* promotion .20* .09 prevention .16 .32*** Note: a .86 -.42*** .02 .75 -.42*** .15 .00 -.08 .16 intrinsic is controlled for extrinsic and vice versa Study 1: intrinsic (F (3; 95) = 3.09, p < .03); extrinsic: (F (3; 95) = 4.94, p < .003). Study 2: intrinsic (F (3; 48) = 4.78, p < .005); extrinsic: (F (3; 48) = 4.69, p < .006). significant at .10 level, * significant at .05 level (one-tailed), ** significant at .01 level (onetailed), and *** significant at the .001 level (one-tailed). In both studies we see the predicted effect: RFQ-proverb promotion is positively related to intrinsic study motivation and not related to extrinsic study motivation, whereas RFQ-proverb prevention is positively related to extrinsic study motivation and not to intrinsic. Indeed, promotion is more strongly related to intrinsic than extrinsic study motivation ( on intrinsic study motivation is .11 higher than extrinsic study motivation in Study 1 and .15 in Study 2), whereas prevention is more strongly related to extrinsic than intrinsic study motivation ( on extrinsic study motivation is .16 higher than intrinsic study motivation in Study 1 and .24 in Study 2). Thus, RFQ-proverb enabled us to differentiate between a chronic promotion and prevention regulatory focus and the differences in study motivations that were related to these regulatory foci. Moreover, we find the same results in the two studies. Replication of the findings gives confidence in RFQ-proverb. USING PROVERBS TO MEASURE REGULATORY FOCUS 53 2.5 Study 3 Study 3 was designed to assess the construct validity of RFQ-proverb in a more unobtrusive way. Several studies of prevention-focus and promotion-focus use response times as indicators. As discussed in the introduction to this chapter, the assumption underlying this method is that the time required to recover a given attribute is a reflection of how readily it comes to mind (Fazio, 1995; Higgins, et al., 1997). It is assumed that accessibility represents activation potential and that stored knowledge with higher activation potential produces faster responses. In other words, a self-representation with a high activation potential produces faster responses than selfrepresentations with a low activation potential. Applying this rationale to RFQ-proverb gives us an interesting opportunity to use our data to ascertain the construct validity in an unobtrusive way. The time that is required to produce an answer to the question: "To what extent does the following proverb reflect all your actions?" is an indication of how readily this answer comes to mind and, thus, of its accessibility (Fazio, 1995). If accessibility represents activation potential, and stored knowledge with higher activation potential produces faster responses, one would expect that questions about proverbs that reflect all a respondent’s actions will be answered more rapidly than those that do not apply. In other words, if RFQ-proverb measures regulatory focus, someone with a chronic promotion regulatory focus should more rapidly respond to promotion proverbs than someone with a chronic prevention regulatory focus, whereas someone with a chronic prevention regulatory focus should more rapidly respond to prevention proverbs than someone with a chronic promotion regulatory focus. 2.5.1 Methodology Study 3. Participants: One hundred and eighty-six students from the Faculty of Social Sciences at the Vrije Universiteit, Amsterdam, participated in an Internet experiment. As thanks for their cooperation, participants received raffle tickets for a mobile telephone. Procedure: The students were made aware of the study via an e-mail message, and there where several announcements during lectures. They could conduct the experiment at either the university or any other place where they had access to the Internet. The participants were given the instructions for RFQ-proverb after a general introduction. The instructions were exactly the same as in Study 1 and Study 2. The participants responded to the 15 proverbs representing a promotion-focus and the 15 representing a prevention-focus one by one alternately. Presenting the proverbs one by one enabled us to measure the time the participants needed to respond to 54 CHAPTER 2 every single proverb. The response times where subjected to a logarithmic transformation to reduce positive skew (see Fazio, 1995) and controlled for the total number of words in the proverbs (promotion 73 and prevention 95). Thus, after administering RFQ-proverb we had two sets of data relating to the same questionnaire, namely, the answers about the extent to which a proverb reflected all their actions, and the time it took to answer the question relating to that specific proverb. If RFQ-proverb measures regulatory focus, the answers and the response time should be negatively related to each other. Note, all the correlations and regressions presented in this section are controlled for overall response time. 2.5.2 Results and Discussion Study 3. Preliminary analyses Table 2.8 provides an overview of the preliminary analyses. As indicated, the correlations with the reaction time on prevention- and promotion proverbs are controlled for overall reaction time. The interrelations between study motivation and regulatory focus are in the expected directions. The RFQ-proverb promotion score is negatively (but not significantly) related to promotion response time and not to prevention response time. The RFQ-proverb prevention score is negatively (but not significantly) related to prevention response time and not to promotion response time. Table 2.8 (Partial) correlations, means, and standard deviations for RFQ-proverb/promotion, prevention score, and response time. 1. promotion score 1 2 3 -- .08 -.08 .02 -- .02 -.14 2. prevention score 3. promotion resp. timea -- 4. prevention resp. timea Note: a 4 .29** -- M 5.17 4.24 11.25 11.23 SD 0.58 0.66 0.41 0.42 transformed using a natural logarithmic transformation and controlled for number of words ** significant at .01 level (two-tailed) USING PROVERBS TO MEASURE REGULATORY FOCUS 55 Main analyses Two hierarchical regression analyses were performed. The first analysis involved the relation of the RFQ-proverb prevention score with the RFQ-proverb prevention and promotion response time controlled for overall response time and the RFQ-proverb promotion score. The second analysis involved the relation of the RFQ-proverb promotion score with the RFQ-proverb promotion and prevention response time, controlled for overall response time and the RFQproverb prevention score. The overall response time was added as control variable to control for a “general response time”, whereas the “alternative type of score” score was added as control variable to control for a general response tendency and a correlation between the two regulatory foci (as suggested by Higgins et al., 1997). The response times where subjected to a logarithmic transformation to reduce positive skew and controlled for the total number of words in the proverbs (promotion 73 and prevention 95). Table 2.9 gives an overview of the results on the prediction of the RFQ-proverb prevention response time. Table 2.9 Results of hierarchical regression analysis with RFQ-proverb/prevention score as outcome variable and prevention and promotion response time as predictors controlled for RFQ-proverb/promotion score and overall response time. R2 F Model 1: overall response time .03 .12 -.03 .01 1.52 -.04 Model 2: overall response time .09 score promotion Model 3: overall response time .01 score promotion 2.67 .05 .12 response time prevention -.21 response time promotion .11 R2 total = .10 F total (4; 181) = 2.01 Note: significant at the 0.10 level (two-tailed) 56 CHAPTER 2 The complete model of the RFQ-proverb prevention score as outcome variable predicted by RFQ-proverb prevention and RFQ-proverb prevention response time, controlled for overall response time and score, explained 3% of the variance (F (4; 181) = 2.01, p = .09). The overall response time ( =. 05, ns) and promotion score ( =. 12, ns) were not statistically significant in relation to the prevention score. More important for the accessibility hypothesis is the RFQproverb promotion and prevention response time. The RFQ-proverb prevention score and promotion response time were statistically not related ( =. 10, ns), whereas the RFQ-proverb prevention score was (marginally) negatively related to the RFQ-proverb prevention response time ( = -.21, p < .10). In other words, the stronger respondents’ prevention regulatory focus, the faster they answer the question about the extent to which prevention proverbs reflect all their actions. Table 2.10 Results of hierarchical regression analysis with RFQ-proverb/promotion score as outcome variable and promotion and prevention response time as predictors controlled for RFQ-proverb/prevention score and overall response time. R2 F Model 1: overall response time .03 4.90* .16* .01 1.52 .16* Model 2: overall response time score prevention .09 Model 3: overall response time .06 6.38*** -.05 score prevention .11 response time prevention .46*** response time promotion -.21 R2 total = .10 F total (4; 181) = 4.89 Note: significant at the 0.10 level (two-tailed), *significant at the 0.05 level (two-tailed), *** significant at the .001 level (two-tailed) USING PROVERBS TO MEASURE REGULATORY FOCUS 57 Table 2.10 provides an overview of the analysis with the RFQ-proverb promotion score as outcome variable predicted by the RFQ-proverb promotion and prevention response time and controlled for overall response time and score. This complete model explains 10% of the variance (F (4; 181) = 4.89, p < .001). Overall response time ( = -.05, ns) and prevention score ( = .11, ns) are not statistically significant in relation to promotion score. The accessibility hypothesis is confirmed in this model as well: the RFQ-proverb promotion response time was (marginally) negatively related to the RFQ-proverb promotion score ( = -.21, p < .10), whereas prevention response time was positively related to promotion score ( = .46, p < .001). Thus, the stronger respondents’ promotion regulatory focus, the faster they respond to promotion proverbs and the slower they respond to prevention proverbs. 2.6 Conclusions Study 3 was designed to test the relation between regulatory focus and response time to the subscales. The stronger the prevention-focus, the faster respondents react to prevention proverbs and the slower to promotion proverbs (though this result was not statistically significant), whereas the stronger the promotion-focus, the faster respondents react to promotion proverbs and the slower to prevention proverbs. If the response time is a reflection of how readily an answer comes to mind and, thus, of its accessibility (Fazio, 1995), we may conclude that a focus with a high activation potential—thus often activated—produces faster responses to proverbs tapping into that construct than proverbs tapping into a focus with low activation potential. The finding that increasing regulatory focus evokes shorter reaction times for the respective subscale and longer reaction times for the alternative subscale may be a result of regulatory inconsistent information. It appears that regulatory consistent information is processed faster, whereas regulatory inconsistent information is processed more slowly. This finding may be due to regulatory fit (Higgins, 1997; 1998) and information sensitivity (Cesario et al., 2004; Higgins, 2000; 2002). Regulatory consistent information seems familiar and evokes fast response times, whereas regulatory inconsistent information seems odd and slows down the response time. However, the most interesting finding for the construct validity of RFQ-proverb is that someone with a chronic promotion regulatory focus measured using RFQ-proverb responds more rapidly to promotion proverbs than someone with a chronic prevention regulatory focus, and vice versa for someone with a chronic prevention-focus. Thus, the scores of RFQ-proverb are negatively related to the response times of these scores. This is exactly what we would expect if RFQ-proverb measures regulatory focus; therefore we assume that the instrument scores tap into CHAPTER 2 58 the regulatory foci constructs. These findings of the response time study have given us, in another—unobtrusive—way, more faith in the construct validity of RFQ-proverb. Taken together, the results of the three studies are promising. The scale characteristics are good (distribution and internal consistency). All factor analyses show two independent factors with high eigenvalues, explaining satisfying variance and items loading in the expected manner. These are important tenets, evidencing the stability of the instrument, particularly as the methods of the studies differed (paper-pencil, and Internet). Despite these different methods, RFQ-proverb elicited data with identical distribution patterns. The fact that RFQ-proverb proved to be very strongly related to RFQ Lockwood indicates that RFQ-proverblike RFQ Lockwoodmeasures regulatory focus. Thus the convergent validity of RFQ-proverb is high. Relating RFQ-proverb to instruments like Personal Need for Structure, Preference for Consistency and the Schwartz Value Scale produced the expected correlations, and this gives confidence in the construct validity. Confidence in construct validity was further enhanced by the results from the response time study. On theoretical grounds we expected a differentiation in study motivation relating to the two regulatory foci. If RFQ-proverb really measures regulatory focus, the instrument must be able to reveal the expected differences in study motivation relating to the foci. Study 1 showed that RFQ-proverb had the ability to distinguish between study motivations in the expected way. Moreover, in Study 2 these findings were replicated. Indeed, intrinsic and extrinsic study motivation is one of the many constructs that we could relate to regulatory focus. However, for the validation of the instrument it is important that the instrument enabled us to differentiate between intrinsically and extrinsically motivated respondents. In other words, Study 1 and Study 2 support the predictive value of RFQ-proverb. Our studies indicate that proverbs are useful to capture the regulatory focus construct. As indicated, proverbs often represent contradictory ways of construing situations and as such appeared to be fruitful to capture the notion of what the often antagonistic promotion- or prevention regulatory motivational states imply. As such, proverbs seem to be useful reflections of norms, values and goals related to regulatory focus. Interesting, recently, we see other researchers starting to use proverbs in self-regulatory studies too (see Briley, Morris, Simonson, 2005). Yet, all that glitters is not gold. One of the problems with proverbs is the comprehension of proverbs by non-native speakers. Indeed, although proverbs and sayings proved to be particularly useful for our instrument because they can say something important with a minimum of text, are linguistically organized to promote memory, and appeared to be a source of predictable differentiation between USING PROVERBS TO MEASURE REGULATORY FOCUS 59 self-regulatory strategies these core qualities might be a pitfall for non-native speakers. Proverbs relate to cultural rules and cultural patterns, which are often implicitly rather than fully explicitly patterns and as such carry subtle nuances as well as social meanings attached to it. Therefore, although comprehension of all kinds of questionnaires might be a major problem for non-native speakers and thus for the validity of the findings, an instrument composed of proverbs might be a little bit more sensitive to this problem. Assessing regulatory focus with a questionnaire, researchers don’t know if they tap into people’s chronic or situational induced regulatory state. Yet we are reasonable convinced that we tap into the chronic regulatory focus. First, we ask to what extent the proverbs apply to all their actions, hence, referring to a more general chronic state. Moreover, the findings in Study 3 reveal that people who score higher on prevention proverbs are faster on prevention than promotion proverbs and vice versa for promotion score, indicating that the dominant chronic focus was more accessible. This validates our argument that we measure chronic regulatory focus. However, these are just assumptions, in future research the intra-individual results on for example a Strength-of-Guide measure (Higgins et al., 1997) could be compared with the score on RFQ-proverb. All in all, RFQ-proverb appears to be a reliable and valid instrument. Moreover, RFQproverb is an unobtrusive measure applicable to all sort of studies, independent of setting and population, where regulatory focus is the concept to be explained. Indeed, we have enough confidence in the instrument to employ it in the studies relating regulatory focus theory to protest participation motives. 60 Chapter 3. Methodology. The main goal of this dissertation is to develop a theoretical model of political protest participation that speaks to questions like who protests in what social movement context and why. Developing a theoretical model is one thing, empirical data confirming the theoretical model is another. How does one collect valid and reliable data on protest behaviour? Data collection on protest participation is rarely comprehensive and, to avoid problems of reliability and validity of the findings, researchers studying collective action behaviour typically apply two methods. The first method studies potential or actual protest behaviour. Based on representative samples from the population, such as the World Value Survey (WVS) or the European Social Survey (ESS), researchers assess who has participated in collective action (cf, Norris, 2002; Inglehart & Catterber, 2002; Welzel, Inglehart & Deutch, 2005). General population surveys have the advantage of sampling people that both did and did not participate in political protest. However, only a small number of citizens actually participate in political protest (in The Netherlands 9% in 2000 and 14% in 2002, Social and Cultural Planning Office, 2004); this implies that many people in the sample may be ignorant about issues relating to political protest participation, not to mention their being motivated to participate in political protest. In the methods relying on general population surveys we see some questions about actual participation, but more often respondents are asked about their intentions to participate in political protest. However, intention to participate is a weak indicator of actual participation (Topf, 1995, cited by van Aelst, Walgrave & Decoster, 2000). Indeed, as social psychologists we know that there is a significant discrepancy between what people want to do and what people eventually do. In their study on the peace movement, Klandermans and Oegema (1987) show that this discrepancy between intention and behaviour also applies to protest behaviour. They report that 60% of the individuals who were motivated to take part in a demonstration organized by the peace movement eventually did not participate. These figures show that “the intention to participate is by no means a sufficient condition” (Klandermans & Oegema, 1987, p. 529) for actual participation. Moreover, the figures on intentions to participate and actual participation in political protest in The Netherlands, mentioned in Chapter 1, reveal the same discrepancy (intentions in 2000 51% and 2002 55%, actual participation 9% in 2000 and 14% in 2002, Social and Cultural Planning Office, 2004, p. 45). The discrepancy between intentions and actual behaviour is not a problem if the variance in intentions is the subject of the study. CHAPTER 3 61 Another often-applied method is protest event studies with an emphasis on events rather than individuals, organizations or movements as units of analysis. Protest event analysis has been developed to systematically map, analyze and interpret the occurrence and properties of large numbers of protests by means of content analysis, using sources such as newspaper reports. Although this line of research produces useful insights into effective collective actions during a certain period (Rucht, Koopmans & Neidhardt, 1998), it is criticized for being subjective and incomplete, because most event data come from newspapers or other news archives, revealing a selection bias problem (van Aelst et al., 2000). The selection bias problem involves assessing the extent to which news sources represent some “true” account of the underlying protest events. Recent studies that use police archives as independent sources of information have generated more insight into the biases of newspaper-based studies, showing clear evidence of selection bias (McCarthy, McPhail & Smith, 1996). Event size was the most important factor in determining whether events were covered. Additionally, they found that some event forms were less likely to be covered and that the amount of news coverage an issue had been receiving predicted the probability that a protest about it would be covered. Protest event analyses are less suitable for the study of the motivational mechanisms behind political protest participation, because at best they offer aggregated information on motivational mechanisms. If one wants to know who participates in protest, why and in what social movement context, it is preferable to ask participants themselves using a protest survey. Protest surveys are written and postal questionnaires distributed among protesters at protest events. Collecting data during protest events implies that the people interviewed are actually performing protest behaviour in contrast to people who intend to perform protest behaviour. In other words, participants are interviewed in the context in which they perform this behaviour, in the heat of the battle, right on the spot. Thus, protest itself becomes the crux of the research (Walgrave, 2005). However, applying this method implies committing the methodological sin of selecting the dependent variable. We will reflect on this issue in the general discussion. Moreover, a demonstration is a living thing and it is tricky to take a sample survey from a boiling mass of people (Walgrave, 2005). We conducted two surveys in two different town squares at protests organized by two different movements at exactly the same time against the same budget cuts proposed by the same government. This Most Similar Systems Design is, from a social psychological point of view, an interesting situation because the organizations emphasized differing aspects of the proposed government policies. Indeed, their interpretation of “what’s going on?”, “who is to blame?” and “how are we going to solve it?”, in other words, their collective action frames, differed. This gave METHODOLOGY 62 us the opportunity to study the influence of social movement context on motivational patterns of individual protesters. This kind of field research implies that it is conducted in a crowded, unpredictable and erratic environment, contrary to the controlled studies conducted in the laboratory or studies based on secondary data sets like general population surveys and newspapers. How can one guarantee reliability, validity and generalizability of the findings? Two Belgians—Stefaan Walgrave and Peter van Aelst—fine-tuned a method to collect data during protest events, and our data were collected using their method. Although obtaining data by using a protest survey is not new, the systematic application proposed by Stefaan Walgrave and Peter van Aelst is. Therefore we will outline its basic principles and then elaborate on data collection during the two demonstrations and the characteristics of the participants in the different studies. However, first we will elaborate on the design of our research. One strategy to study the relative importance of the different participation motives contingent on social movement context is to compare its relative importance over different protest events, in other words, a comparative research design. 3.1 Comparative research designs within protest participation research. Comparative designs may incorporate comparisons of movements or events or other features of social movements across space or time. Such comparisons are rare (Klandermans & Smith, 2002). Yet, as Klandermans and Smith hold, “comparative research of movement participation is important. It tells us that what holds for a participant in one movement, or at one point in time, or at one place is not necessarily true for a participant in another movement, or at a different time or place” (p.6). Comparisons across space examine the same movement in different locations and are the most common. A classic example is Walsh’s (1988) study of citizens and activists in four communities in the neighbourhood of Three-Mile Island, and a more recent illustration is a study wherein ideas stemming from the American women’s movement are embraced and incorporated in, respectively, the Dutch and the Spanish women’s movement. A comparison of the implementation of the same ideas shows that the diffusion and adaptation process is context dependent (Roggeband, 2002). Each study demonstrates that the dynamics of participation are shaped by characteristics of the local communities in which the movements are embedded. Had these authors neglected to make these comparisons (either by restricting themselves to a single community or by simply analyzing aggregated data), we would erroneously believe that the dynamics of participation in each community were the same. Such comparisons are important CHAPTER 3 63 because they may reveal diverging political, economic or social psychological dynamics of movement participation. Comparisons across time examine the same movement over a certain time span. Movements expand and contract in phases of mobilization and demobilization; these waves of expansion and contraction can be analyzed by comparison across time. A good example is the longitudinal study of the farmers’ movement by De Weerd and Klandermans (1999). Most studies on identification with the group at stake and protest participation on behalf of the group are basically correlational. Thus, identification can cause participation, but it could be the other way round as well. In their field study with a comparative research design across time, De Weerd and Klandermans were able to disentangle this causality pitfall. These scholars demonstrated that identification with farmers stimulates participation in protest on behalf of farmers and, subsequently, participation increases identification with the farmers.! A comparison across movements or protest events enables us to answer different questions. The most common question in a comparison of movements concerns the similarities and differences between participants in different movements or protest events. These differences may concern demographic characteristics, motivational dynamics, identity, attitudes and ideology. An example is Klandermans’ (1993) comparison of participants in the labour, peace and women’s movements in The Netherlands. Drawing on the aforementioned distinction in action orientations, the study tests the hypothesis that each movement appealed to different action orientations: a power orientation in the case of the labour movement, a value orientation in the case of the peace movement and a participation orientation in the case of the women’s movement. Yet another example of comparing protest events is the comparison by van Aelst et al., (2000) of four national mass demonstrations in Brussels. In this study, the authors compare participants from the different protest events amongst both the demonstrations and “the average citizen” of Belgium. With this strategy they were able to answer questions about similarities and differences between the different protest events and questions concerning similarities and differences between protesters and “average citizens”. All in all, these examples show the advantage of comparative research designs and the kind of questions that can be answered using them. However, we are not aware of studies comparing the relative weight of motivational mechanisms instigating protest participation over different protest themes. If we want to achieve an increased theoretical understanding of motivational mechanisms fuelling protest behaviour, we need comparisons; but the comparisons must be methodologically sound. According to Klandermans and Smith (2002), a key issue to ensure methodological soundness is that both the sampling frames and the questionnaires are METHODOLOGY 64 comparable. Firstly, the samples: “In order to draw equivalent samples one needs comparable sampling frames of the groups one wants to compare […] after all, we want to be able to attribute the differences we find between two or more samples to real contextual differences rather than to sampling biases” (p. 10). Secondly, the questionnaires: “That is, questions must not only have comparable wording, but they must also have the same meaning for each group in a study (Klandermans & Smith, 2002, p. 10, italics original). In the studies reported in this thesis, we applied the sampling method proposed by Walgrave and colleagues (van Aelst & Walgrave, 2001; Van Aelst et al., 2000) to give every protester an equal chance to be approached with the request to fill in a questionnaire. In our view this is the most sophisticated method to ensure comparable samples. 3.3 Protest surveys. As indicated, interviewing participants at protest demonstrations is not new, but interviewing participants in a systematic manner is a relatively new technique. Favre, Fillieule and Mayer (1997, cited by van Aelst & Walgrave, 2001, p. 468 speak of “a curious lacuna in the sociology of mobilization”. Van Aelst and Walgrave (2001) found only a few studies applying this technique. In 1979, Ladd, Hood and van Liere (1983, cited by van Aelst & Walgrave, 2001) conducted interviews at a large anti-nuclear demonstration in Washington. Their objective was to identify the extent to which participants share common positions on ideological issues. When demonstrators in Sheffield took to the streets in 1983 to protest against the visit of Mrs Thatcher, the British Prime Minister, Waddington, Jones and Critcher (1988, cited by van Aelst & Walgrave, 2001) conducted 300 interviews to document the socio-demographic profile of the protesters. However, neither study gave much explanation about how the survey was set up and administered. Waddington stated: “Our survey of the demonstrators, which was random in the literal rather than the scientific sense, provided a rough profile of the demonstrators” (Waddington et al., 1988, p. 473, cited by van Aelst & Walgrave, 2001). At the beginning of 1994, Favre et al. (1997) carried out three surveys at large protest marches in France. They designed a strategy to give all participants an equal opportunity of being interviewed. The method of Favre and colleagues was refined by Walgrave and van Aelst (van Aelst et al., 2000). Walgrave and van Aelst refined Favre et al.’s method in the context of several Belgian protest marches. Given the unpredictable nature of protest action, representativeness is still the main obstacle as far as this method is concerned. Walgrave and colleagues proposed two techniques to increase the degree of reliability. The first proposed technique is a method to CHAPTER 3 65 guarantee every protester an equal chance to participate in the study. In the protest events studied by Walgrave and van Aelst, a dozen interviewers distributed approximately 700 questionnaires during the actual protest march itself, while a number of “reference persons” ensured that the same number of rows was skipped throughout. On this basis, at every relevant row a protester was asked if he or she was willing to fill in the questionnaire. The respondents were asked to fill in the questionnaire at home and to mail it to the researchers. However, this method is not always applicable, as Boekkooi (2005) shows. In her study, the protest event (anti-war demonstration, 2003) took place in a square in Amsterdam rather than being a march on a broad avenue in Brussels. To fit the environmental circumstances of a protest event taking place in a square rather than on a broad avenue, the method of Walgrave and colleagues was slightly modified. To give every protester an equal chance to participate in the study, interviewers were equally distributed around the square on the outer edge of the protest event. The interviewers were instructed to distribute one questionnaire to a protester on the outer circle, followed by another, ten steps inwards, and so on until the centre of the circle was reached. Although the studies in Brussels and Amsterdam are slightly adapted to different environmental circumstances, the basic assumption underlying both methods is the same. To ensure a reasonable degree of representativeness of the sample of addressed protesters, researchers have to agree on a device that guarantees that all protesters spread around the area where the protest event is taking place have the same chance of being addressed by one of the interviewers. The second technique proposed by Walgrave and colleagues to enhance the representativeness of the findings is to conduct face-to-face interviews. In addition to the mail surveys, a much smaller sample of protesters was questioned orally before the protest set off. Each interviewer posed a few questions at random to about ten waiting protesters about the main predictor variables and some demographics. These short, face-to-face interviews were used primarily to evaluate the representativeness of the mail survey. Experience shows a response rate of around 100% to face-to-face interviews. Thus, provided proper sampling is undertaken, these face-to-face interviews can serve as check for the question as to whether the surveys are biased (due to non-response). A non-response bias may arise due to the fact that people who respond to surveys answer questions differently than those who do not (Benson, Booman & Clark, 1951; Gough & Hall, 1977, cited by Walonick, 1997). Despite the fact that the response rates of the surveys are very satisfactory for a mail survey with no reminders (40-45%), nothing is known about the protesters who did not return their questionnaires. Some degree of reliability (van Aelst & Walgrave, 2001) is suggested by the fact that hardly anyone refuses a face-to-face interview METHODOLOGY 66 and that there are—in general—no significant differences between responses from the two types of interviews. Walgrave and colleagues agree that representativeness is still the main obstacle as far as this method is concerned. “This is probably also why very few researchers have used the technique” (van Aelst & Walgrave, 2001, p. 474). However, they continue, “Our experience shows that with sound preparation and sufficient interviewers, a representative picture of the protesters can be obtained” (van Aelst & Walgrave, 2001, p. 474). We conducted protest surveys, using the two techniques proposed by Walgrave and colleagues, during a demonstration organized by two social movement organizations, the trade unions and the Turn the Tide alliance (an alliance against neo-liberal policies encompassing 500 organizations representing 500,000 members). Our studies were designed to collect data with respect to motivational mechanisms leading to protest participation compared over different social movement contexts. With the discussed theoretical frame of protest participation as the point of departure, we constructed several measures that would help to explain why people were motivated to participate in one protest event rather than the other. In the following sections we elaborate on the procedures and characteristics of the participants of both studies. The specific measures relating to the motivational mechanisms are described in the method sections of the specific chapters. 3.4 Procedure Study 4 and Study 5. We conducted protest surveys, using the two techniques proposed by Walgrave and colleagues, during a demonstration organized by two social movement organizations, the trade unions and the Turn the Tide alliance (an alliance against neo-liberal policies encompassing 500 organizations representing 500,000 members). Our studies were designed to collect data with respect to motivational mechanisms leading to protest participation compared over different social movement contexts. With the discussed theoretical frame of protest participation as the point of departure, we constructed several measures that would help to explain why people were motivated to participate in one protest event rather than the other. In the following sections we elaborate on the procedures and characteristics of the participants of both studies. The specific measures relating to the motivational mechanisms are described in the method sections of the specific chapters. In Study 4 (the union) and Study 5 (Turn the Tide) we distributed postal surveys and conducted short, face-to-face interviews during the demonstration to evaluate the CHAPTER 3 67 representativeness of the mail survey. As we were interested in the differences in motivational mechanisms between the two mobilizing channels, we treated the two demonstrations as two separate populations. Consequently, 123 (Turn the Tide) and 115 (union) face-to-face interviews of about ten minutes were conducted and 500 x 2 postal survey questionnaires were distributed. Following the same procedure as Walgrave and colleagues, we started with the face-to-face interviews before the programme of invited speakers began. The interviewers were equally distributed around the two squares and they were instructed to interview one protester every ten steps to the centre of the square. Practically no protester refused to participate in the face-to-face interview. When the programme of invited speakers started, the distribution of the 500 survey questionnaires began in both squares. In line with exactly the same procedure, the interviewers were equally distributed around the two squares, and at every ten steps to the centre protesters were asked if they were willing to participate in a survey. They were asked to fill in the questionnaire at home and then post it to us (postage pre-paid by us). The response rate was very satisfying for a postal survey without reminders: the overall response rate was 44% (unions 233: response rate 47% and Turn the Tide 209: response rate 42%). 3.5 Participants Study 4 and Study 5. The two studies were designed to test our theoretical framework of protest participation. Study 4 and Study 5 represent two mobilizing channels with somewhat different goals demonstrating at the same time in the same city but in different squares. Our empirical material consists of two separate samples wherein in total 238 face-to-face interviews (115 in Study 4 and 123 in Study 5) were conducted and 1,000 questionnaires were handed out, of which 442 questionnaires (233 in Study 4 and 209 in Study 5) were returned. First, we will describe the characteristics of the participants of both samples based on the postal questionnaires. After that, we will elaborate on the differences between the face-to-face interviews and the postal questionnaires, in other words, the non-response bias. Table 3.1 provides an overview of the characteristics of the protesters at both protest events. METHODOLOGY 68 Table 3.1 Characteristics (survey) of the protesters of the different protest events. Variables Study 4 (union) Study 5 (alliance) N= 233 209 Gender (% males) 56% 48% Age: mean 52 years 44 years % (post) highschool 68% 45% % college (Bsc/Msc) 29% 53% non-member 46 (19%) 91 (44%) member 194 (81%) 117 (56%) Educational level: Relation to organization: Of the union sample (N = 233) 56% were male, with an average age of 52 years, and moderately educated (68% had finished high school and 29% had a degree). Eighty-one percent of the participants were union members. The Turn the Tide sample (N = 209) was 48% male, with an average age of 44 years, and highly educated (45% had finished high school and 53 % had a degree). Fifty-six percent were members of an organization affiliated to the Turn the Tide alliance. Social movement organizations with somewhat different goals seem to attract somewhat different protesters. First, although the Turn the Tide sample shows male/female ratios of around 50%, males are overrepresented in the union sample (56%). Van Aelst et al. (2000) indicate that the male/female ratio depends on the theme of the protest event. Interestingly, in a demonstration with social security as its theme, van Aelst et al. (2000) report the same overrepresentation of males (65%). Next, compared to the union, the majority of Turn the Tide protesters tend to be relative young (Turn the Tide M = 44 years, union M = 52 years) and well educated (Turn the Tide 53% with a degree, union 29% with a degree), in other words, the protesters of Turn the Tide, tend to have a social structural profile similar to the presumed base of support of so-called new social movements (Scott, 1990). In addition to providing the traditional socio-economic variables, these surveys enabled us to zoom in on the context of the mobilization. Since organizations are still widely regarded as essential for successfully mobilizing protesters (McAdam, McCarthy & Zald, 1996), we expected CHAPTER 3 69 to see a higher number of association members among the protesters. The figures confirm this: protesters are members of an association (Turn the Tide 56%) or union (union 81%). Protesters are more likely to be members of an association or union than the average Dutchman (in 1999, 34% of Dutch citizens over 18 years of age were members of at least one social organization (the Amenities and Services Utilisation Survey, scp-avo, 1999). Hence, compared to the union, the Turn the Tide protest seemed to be able to mobilize more non-members. We will return to this in the general discussion. Thus, the union and Turn the Tide samples differ, and these differences seem to relate to the fact that different mobilizing contexts attract different protesters. Yet, so far, nothing is known about the demonstrators who did not return their questionnaires. This brings us to the issue of the representativeness of our samples. 3.6 Representativeness of the samples of Study 4 and Study 5. As indicated, the response rate was very satisfying for a postal survey without reminders. Walgrave and colleagues report response rates of the same magnitude. Response rate is seen as the single most important indicator of how much confidence can be placed in the results of a mail survey (Walonick, 1997). Yet, comparing the face-to-face interviews with the postal questionnaires offers us the opportunity to raise a corner of the veil on the black hole of nonrespondents. A dataset is representative if there is no difference between respondents and nonrespondents. In the context of representativeness, the extent of the comparability of the demographic characteristics and predictor variables of the face-to-face interviews and the postal questionnaires is most important. As indicated, the fact that hardly anyone refused a face-to-face interview makes this dataset an anchor in testing the representativeness of the sample. Following Walgrave and colleagues, we compared the demographic characteristics and the main predictor variables (instrumental, identity, ideology and group-based anger motives) of the postal surveys and the face-to-face interviews to evaluate the representativeness of our samples. First we will elaborate on the representativeness of the samples derived from the demographic variables, followed by the motivational variables. Table 3.2 provides an overview of the demographic characteristics of the respondents and the most important predictors in Study 4 and Study 5 of the face-to-face interviews and the surveys. METHODOLOGY 70 Table 3.2 Characteristics (survey and interviews) of the protesters of the different protest events. Study 4 (union) Study 5 (alliance) Variables interviews survey interviews survey N= 115 233 123 209 Gender (% males) 61 % 56 % 50 % 48 % Age: mean 51 years 52 years 45 years 44 years - % (post) highschool 64 % 68 % 47 % 45 % - % college (Bsc/Msc) 31 % 29 % 48 % 53 % non-member 27 (23%) 46 (19%) 46 (38%) 91 (44%) member 88 (77%) 194 (81%) 77 (62%) 117 (56%) 4.70 4.51 4.40 Educational level: Relation to organization: Main predictors **: Instrumental motives a b 4.38 a 2.69c Identity motives 4.76 Ideology motives 6.53a 6.18b 6.63a 6.25b Group-based anger 5.01a 5.45c 5.67a 5.21c Note: 3.36 3.51 * on a Likert scale with 1 (not at all) to 7 (very much) a-b a-c Means differ significantly from each other at p < .001 level Means differ significantly from each other at p <.01 level We can be brief on the demographic characteristics of the protesters. The characteristics of the interviewed and surveyed protesters did not differ, male and female were equally likely to be interviewed and to return their survey, and neither age nor educational level seems to influence the willingness to return the postal questionnaire. Affiliation with the mobilizing organization seems not to influence the willingness to return the questionnaire; both members and notmembers were equally willing to return the questionnaire. The motivational variables, on the other hand, do differ. Both identity and ideology motives are stronger during the demonstration than after the demonstration for protesters from both the union federation and Turn the Tide. Apparently, the mere presence of members of the shared collective identity provides shared ideological definitions of ambiguous situations and enhances feelings of commitment and solidarity. This is an interesting finding, but might be CHAPTER 3 71 problematic for the representativeness of our samples. Yet, the fact that there is a similar pattern for both the union federations and Turn the Tide—both identity and ideology motives are stronger during the protest event—makes it less problematic. Group-based anger motives, however, show a different pattern; union protesters are angrier at home, whereas Turn the Tide protesters are angrier during the demonstration. This finding may be a combined effect of identity salience and mobilizing context. It may be that, in the context of value-oriented protest where ideology motives prevail, the mere presence of people sharing one’s view increases the anger about such an ambiguous motive as ideology. In the context of power-oriented protest where instrumental motives will prevail, the mere presence of other people may enhance protesters’ perception of the effectiveness of the protest activity and subsequently lower their anger. These are just speculations; when we are discussing the findings regarding group-based anger we will take this finding into consideration and we will return to this difference in the general discussion. In summary, the comparison between the demographic characteristics of the protesters from the face-to-face interviews and the surveys showed no significant differences. This, in combination with the fact that hardly anyone refused a face-to-face interview, suggests some degree of reliability. Yet, the response pattern of the participation motives differed; identity and ideology motives were stronger during the demonstration than after the demonstration for both the union and Turn the Tide; and unionists were angrier at home, whereas Turn the Tiders were angrier during the demonstration. However, this seems more a reflection of the influence of environment on the answer tendencies than a non-response bias effect. Thus, despite the crowded, unpredictable and erratic setting wherein data were collected, these evaluations give some confidence in the representativeness of the samples. The experiences of both van Aelst and colleagues and ourselves show that “with sound preparation and sufficient interviewers, a representative picture of the demonstrators can be obtained” (van Aelst et al., 2001, p. 474). 72 Chapter 4. Instrumental, identity, ideological, and group-based anger and context: the relative weight of participation motives. We proposed that people may employ four fundamental motives to participate in protest: instrumental, identity, ideology and group-based anger motives. As indicated, we assume that, depending on individual differences and social movement contexts, the relative weight of these motives differs. In this chapter we elaborate on the differences in the importance of the fundamental participation motives employed due to social movement context. In this approach, Turner and Killian’s (1987) concept of action orientation is relevant. The general assumption underlying our hypotheses about the relationship between action orientation and the relative weight of fundamental participation motives is that, depending on their action orientation, movements appeal to different motives. How action orientation influences the relative weight of participation motives is elaborated in the comparison of participation in two protest events, one against the erosion of early-retirement rights, and the other against the neo-liberal policies of the government. Applying the distinction in action orientation proposed by Turner and Killian (1987), we assume that the labour movement demonstration against the erosion of early retirement rights fits the description of action that is predominantly power-oriented. We assume that the protest event of the Turn the Tide alliance, on the other hand, is more value-oriented. By stressing anti-neo-liberal and progressive policies they emphasize the ideology behind their claims, so giving participants the opportunity to express their discontent with proposed government policies. What path to protest participation will prevail, depending on these differences in action orientation? In Chapter 1 we proposed a mediator-moderator model of protest participation. In this chapter we test some hypotheses resulting from the model. We differentiate social movement context in terms of a movement’s action orientation. We hypothesized that power-oriented action tends to appeal to instrumental motives in addition to identity and group-based anger motives. Moreover, we expect identity motives to moderate instrumental motives and instrumental motives to mediate between identity and groupbased anger motives. With regard to value-oriented protest, we proposed, on the other hand, that it tends to trigger ideology motives also, in addition to identity and group-based anger motives. In this social movement context, we assume that identity motives moderate ideology motives and that ideology motives mediate between identity and group-based anger motives. CHAPTER 4 73 In statistical terms, we test in this chapter a mediator-moderator model accounting for protest participation. The assumption underlying the social movement context moderator model is that these two movement contexts differ in emphasizing instrumental or ideological reasons to participate. As a consequence, the moderator (identity motives) moderates different mediators (instrumental or ideology motives). In power-oriented protest, identity motives will moderate instrumental motives, while instrumental motives will mediate between identity and group-based anger. Finally, in value-oriented protest, identity motives will moderate ideology motives and ideology motives will mediate between identity and group-based anger motives. 4.1 Method 4.1.1 Sample To test our hypotheses with respect to action orientation and the relative importance of participation motives we compared data in relation to the participation motives from Study 4 and Study 5 (see Chapter 3, Methodology). 4.1.2 Measures All data for the analyses in this study were taken from postal survey questionnaires and were selfassessments. The main measures fall into the following categories: value and expectancies (instrumental motive), identification with the group (Study 4: union organization, Study 5: organizations falling under Turn the Tide alliance umbrella), ideology motives, group-based anger motive, and motivational strength. All variables were measured on a Likert scale ranging from 1 (not at all) to 7 (very much). Instrumental motive. Following Klandermans (1984) and Simon et al. (1998; 2003), instrumental participation motives were operationalized in terms of values and expectancies. The value of expected outcomes was multiplied by the expectancy that the protest event would contribute to achieving the outcome. The value component was stated as: “To what extent is your personal situation affected by the government plans concerning early retirement rights?” The expectancy component was stated as: “To what extent do you think that this protest event will contribute to persuading the government not to implement their plans concerning early retirement rights?” Identity motive. Respondents’ identification with the union (Study 4) or an organization of the Turn the Tide alliance (Study 5) was measured with the same items referring to the different groups. We measured emotional significance (e.g., “I like being part of this group”), commitment (e.g., “I feel committed to this group) and shared “we” (e.g., “I have much in common with other members”) and involvement (e.g., “I am involved in this group”). In all three samples, the items THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 74 loaded on a single factor and accordingly we calculated one identity motive measure (identity motive: Study 4 = .96, and Study 5 = .98) for each respondent by averaging responses over items. Ideology motive. This motive was measured by six items: “I am protesting because: I am worried about the proposed government policy”, “I am concerned about the proposed government policy”, “I want to take my responsibility”, “The proposed government policy is against my principles”, “I find the proposed government policy unfair”, “I find the proposed government policy unjust”. The “proposed government policy” differs for the two studies: Study 4 concerns early retirement rights, and Study 5, so-called neo-liberal policies. In both samples the items loaded on a single factor and accordingly we calculated one ideology motive measure (ideology motive: Study 4 = .85, and Study 5 = .88) for each respondent by averaging responses over items. Group-based anger. Following van Zomeren et al. (2004), we measured group-based anger with four items derived from Mackie et al. (2000). “Thinking about the government proposals makes me feel…(angry, irritated, furious, displeased). We obtained a reliable scale (Study 4 = .88, and Study 5 = .81). Motivation to participate in the protest. Respondents indicated the strength of their motivation to participate with the following item: “How determined were you to participate in this protest event?” 4.2 Results 4.2.1 Preliminary analyses. Table 4.1 provides the correlation matrices, means and standard deviations of the Study 4 and Study 5 variables. 75 CHAPTER 4 Table 4.1 Correlations, means, and standard deviations for instrumental, identification, and ideological participation motives, and motivation for Study 4 and Study 5. M Study 4 1. Gr.-based anger 5.45 Power-oriented 2. Instrumental♦ Study 5 1.48 23.45 15.96 3. Identity 4.46 4. Ideology 5. Motivation 2 3 4 5 -.28** -- 2.15 .08 .39** -- 6.40 0.74 .34** .20** .07 6.51 0.83 .27** .19** .24** 1. Gr.-based anger 5.21 1.55 -- Value-oriented 2. Instrumental♦ Note: 1 SD -.33** -- 19.12 14.85 .10 -- 3. Identity 3.76 2.72 .16* -.07 -- 4. Ideology 6.43 0.77 .41** .12 .26*** -- 5. Motivation 6.29 1.23 .43** .12 .28** .52** -- Valid N (listwise) Study 4 = 220 and Study 5 = 205 ♦ Value times expectancy (V x E) * p < .05, ** p < .01 The correlations of the participation motives with the strength of motivation in the poweroriented protest event show that instrumentality, identity, ideology and group-based anger motives are positively related to motivation. The pattern of correlations in the context of poweroriented protest shows that instrumental motives are positively related to identity, ideology, group-base anger and strength of motivation, whereas the pattern of correlations in the context of value-oriented protest reveals that instrumental motives are not significantly related to the other motives and strength of motivation. This suggests that instrumental motives are irrelevant in value-oriented protest. Identity motives in the context of power-oriented protest are positively related to instrumental and not to ideology and group-based anger motives. Identity motives in a valueoriented protest event, however, are positively related to ideology and group-based anger motives. 4.2.2 Analyses strategy. Recall that we hypothesized a main effect for instrumental, identity and group-based anger motives for power-oriented protest, and identity, ideology and group-based anger motives for THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 76 value-oriented protest. Next, we are interested in determining whether identity motives moderate the influence of instrumental rather than ideology motives upon motivation to participate in the power-oriented protest event. In contrast, in the context of the value-oriented protest, we expect identity to moderate ideology rather than instrumental motives. Finally, we wish to test the mediating effects of instrumental motives in power-oriented protest events, and of ideology motives in value-oriented protest events, on identity and group-based anger motives. Main effects. To test for replication of the findings of Simon et al. (1998) and van Zomeren et al. (2004), and to investigate the extent to which ideology adds to the explained variance in the strength motivation two separate hierarchical regressions were conducted for both social movement contexts separately (four regressions in total). In the first hierarchical regression the four motives were hierarchical introduced. First the instrumental path suggested by Klandermans (1984) was entered (Model 1), followed by the dual path suggested by Simon et al. (1998, Model 2), and the dual-path proposed by van Zomeren et al. (2004, Model 3) was introduced. In Model 4 the three paths are combined, and finally an ideology path was entered (Model 5). Moderating effects. The strategy commonly used to test moderating effects involves entering the main effect variables in the first step followed by entering the interaction term in the next steps. Interactions are indicated by a significant increase in the R2 statistic. Procedures outlined by Aiken and West (1991) were followed to calculate regression slopes and plots. Following Aiken and West, before the interaction terms were calculated, the criterion variables and all continuous predictor variables (i.e., instrumental, identity, ideology and group-based anger) were standardized. To test whether identity moderates instrumental motives in the powerrather than the value-oriented protest, the interaction of identity motives and instrumental motives was incorporated in the model (model 6). Finally, to test whether identity moderates ideology motives in the value- rather than the power-oriented protest, the interaction of identity motives and ideology motives was incorporated in the model (model 7). Mediating effects. To test whether identity not only moderates what path to protest participation will prevail but also influences what makes group members angry, a set of mediational analyses were conducted. More precisely, we hypothesized that instrumental rather than ideology motives would explain the relation between identity motives and group-based anger in power-oriented protest events, whereas ideology rather than instrumental motives would explain the relation between identity motives and group-based anger in value-oriented protest events. To test our ideas concerning the mediational role of group-relevant appraisals on identity and group-based anger motives, we followed Baron and Kenny’s (1986) three-step procedure. CHAPTER 4 77 First, according to this procedure, the predictors should be significantly related to the mediator variables; second, the predictors should be related to the outcome variables; and, third, the mediating variables should be related to the outcome variables with the predictors included in the equation. If the three conditions hold, at least partial mediation is present (Liden et al., 2000). If the independent variables have non-significant beta weights in the third step, then complete mediation is present (Baron & Kenny, 1986). As a result of the hierarchical regressions and the mediation analyses, a baseline model accounting for motivation to participate in political protest will be proposed. Subsequently, this baseline model will be tested using Structural Equation Modelling, whereafter this baseline model will be modified according to the different social movement contexts. 4.2.3 Main effects. The results of the first analysis (power-oriented) demonstrate that instrumental, identity, groupbased anger and ideology participation motives each have a significant main effect on the strength of motivation to participate in the power-oriented union protest event (see Model 1, 2, 3, and 5, Table 4.2). In the first Model of the power-oriented protest, the findings of Klandermans (1984) are replicated, demonstrating that instrumental participation motives influence the motivation to participate. In the second Model, the findings of Simon et al. (1998) are replicated, showing that instrumental and identification concerns both independently influence the motivation to participate. In the third Model, the findings of Van Zomeren et al. (2004) are replicated, showing instrumental and group-based anger motives both independently influence the strength of motivation to participate in power-oriented protest. Entering instrumental, identity and groupbased anger motives in one model (Model 4) reveals that instrumental motives, net of the other two motives, no longer have a unique influence on the strength of motivation to participate. Interestingly, contrary to our expectations, instrumental motives seem to be mediated by groupbased anger. We will return to this in our mediational analyses. In the fifth Model, our new pathway to protest participation is introduced: ideology motives. The data reveal that identity, group-based anger and ideology motives independently influence the strength of motivation to participate in power-oriented protest. Moreover, the data show that the ideology path is relatively the most important path ( = .24) spurring strength of motivation to participate in a poweroriented protest event, followed by identity ( = .20), group-based anger ( = .15) and finally by instrumental participation motives ( = .05). This indicates that the influence of the ideology path THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 78 on motivation to participate in power-oriented protest events is nearly five times higher than the influence of the instrumental path. Table 4.2 Hierarchical regressions of motivation to participate in power- and value-oriented protest events on instrumental, identity, ideology and group-based anger motives. Motivation to participate Power-oriented Model 1 Model 2 Instrumental (A) .23*** .15* Identity (B) .20*** Anger (C) Model 3 .15* .23*** Model 4 Model 5 Model 6 Model 7 .08 .05 .08 .07 .20*** .20*** .17* .15* .22*** .15* .17* .14* .24*** .24*** .32*** Ideology (D) BxA -.14* BxD -.07 -.32*** Model F 11.30*** 9.49*** 10.78*** 10.13*** 11.19*** 10.08*** 13.54*** df (1,208) (2,207) (2,207) R2 .05 .08 .09 2 R change .03*** .04*** (3,206) .13 a b .05*** (4, 205) (5, 204) (6, 203) .18 .20 .29 .05*** .02* .09*** Value-oriented Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Instrumental (A) .12 .09 .10 .06 .05 .05 .22*** .14* .13* .15* .40*** .23*** .23*** .22*** .45*** .45*** .38*** Identity (B) .14 .28*** Anger (C) .43*** Ideology (D) BxA -.04 BxD -.03 -.14* Model F 2.66 8.67*** 22.77*** 18.66*** 29.23*** 23.37*** 20.53*** df (1,170) (2, 169) (2, 169) (3,168) (4,167) (5, 166) (6,165) R2 .02 .09 .20 .25 .41 .41 .43 .16*** .00 .02* 2 R change Note: .07*** .18*** a .16*** b Coefficients are standardized regression weights (betas) * p < .05, **, p < .001 power-oriented: F total F (6; 209) = 13.64, p < .001, R2 total = .29 value-oriented: F total (6; 165) = 20.53, p < .001, R2 total = .43 a 2 R change between model 1 and model 3 b 2 R change between model 2 and model 4 79 CHAPTER 4 As expected, identity, group-based anger and ideology motives explain the motivation to participate in value-oriented protest events, whereas instrumental motives do not significantly influence the motivation to participate (see Steps 1, 2, 4 and 5, Table 4.2). In the first Model of the value-oriented protest, the findings of Klandermans (1984) are not replicated; indeed, instrumental motives do not influence the strength of motivation in valueoriented protest. The second Model of the value-oriented protest reveals that the findings of Simon et al. (1998) are not replicated either: instrumental motives do not influence the strength of motivation, whereas identity motives do. The findings of van Zomeren et al. (2004) are not replicated either: again, instrumental motives are not related to strength of motivation, whereas group-based anger motives are. Entering instrumental, identity and group-based anger motives in one model (Model 4) reveals that identity and group-based anger have a unique influence on motivational strength in value-oriented protest. In the fifth Model, our new pathway to protest participation is introduced: ideology motives. The data reveal that, in the context of valueoriented protest, identity, group-based anger and ideology motives independently influence the strength of motivation to participate. Ideology motives ( = .45) are a strong predictor of motivational strength in the value-oriented protest event, followed by group-based anger motives ( = .23) and identity motives ( = .14). The influence of the ideology path on motivation to participate in a value-oriented protest event is more than three times greater than the influence of identity motives on motivation to participate; this supports our assumption that people participate in a value-oriented action to express their view. Additionally, although ideology motives influence strength of motivation to participate in power-oriented protest events, these studies show that the influence of the ideology path on motivation to participate in value-oriented protest events ( =. 45) is much stronger than in power-oriented protest events ( = .24). 4.2.4 Moderating effects of identity and social movement context. We also formulated hypotheses on the interaction between identity and instrumental motives and identity and ideology motives (see Model 6 and 7, Table 4.2). We expected that identity motives would moderate instrumental and ideology motives. Following Aiken and West (1991), interactions are indicated by a significant increase in R2 if the interaction term is entered in the regression. In the context of power-oriented protest, identity motives moderate, as expected, instrumental motives; yet, contrary to our expectations identity motives moderate ideology motives as well. In the context of value-oriented protest, only ideology motives are moderated by identity motives. THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES We hypothesized that the 80 relationship between motivational strength and instrumental/ideology motives was contingent not only upon identity, but also upon social movement context. In the abovementioned analyses, context was not taken into account as a variable, although the influence of context was inferred from the difference between the powerand the value-oriented protest. So, to test our hypothesis concerning the moderator model of social movement context more precisely, we conducted a hierarchical regression on the two samples together with two three-way-interaction terms: social movement context (“0” = alliance, “1” = union), identity, and instrumentality, and social movement context, identity, and ideology. To test this social movement context as moderator hypothesis, we first regressed motivational strength on instrumental, ideology, identity and context. In a second step, the products of each variable were entered in the regression. The third step contains the product of all four predictor variables. It is important for the social movement and identity as moderator hypothesis that both 3-way interactions were significant: the identity/instrumental/context: F (5; 376) = 31.63, p < .001, .001, = -.08, p = .06; the identity/ideology/context: F (6; 376) = 32.06, p < = -.22, p < .001, indicating that the relationship between instrumental and ideology motives and motivational strength varies across levels of identity, mobilizing context, and/or a combination of identity and mobilizing context. To interpret these findings we have plotted the relationship between motivational strength and instrumental and ideology motives at high and low levels of identity for power- and valueoriented protest separately (Aiken & West, 1991). First, each predictor was standardized. For the outcome variable (motivational strength), interaction effects were plotted using the parameter estimates derived from the standardized regression equations. Predicted values were computed using scores that were one standard deviation below and above the mean of instrumental and ideology motives (for low and high identity motives, respectively, Aiken & West, 1991). In Figure 4.1, graphs a and b display the influence of instrumental and ideology motives on motivational strength for low and high identifiers in the context of power-oriented protest. Graph c of Figure 4.1 displays the influence of ideology motives on motivational strength for low and high identifiers in the context of value-oriented protest6. 6 The moderating effects of identity on instrumental motives in the context of value-oriented protest reached no significance ( = -.09, p = .20). Therefore this interaction is not plotted. 81 CHAPTER 4 identity low identity high identity low identity high identity high 1 1 -1 motivation motivation motivation 1 identity low -1 - 1 sd +1sd instrumental power-oriented a -1 - 1 sd +1sd - 1 sd ideology power-oriented b +1sd ideology value-oriented c Figure 4.1. Motivational strength as a function of the interaction of identity and instrumental motives and identity and ideology motives. The significant interaction in the first graph on identity and instrumental motives reflects the fact that the motivational strength of weak identifiers in the context of power-oriented protest increases as the strength of their instrumental motives increases (simple slope = .27, p < .001), whereas the motivational strength of strong identifiers remains invariably high (simple slope = .03, ns), irrespective of the level of instrumental motives. Apparently, the less protesters identify with the union, the more their motivational strength is influenced by instrumental motives, whereas the motivational strength of strong identifiers is high, irrespective of the level of instrumental motives. The significant interaction of the second graph, the interaction of identity and ideology in the context of power-oriented protest, reflects the same pattern: the motivational strength of weak identifiers is strongly influenced by their ideology motive (simple slope = .63, p < .001), whereas the motivational strength of strong identifiers is unaffectedly high (simple slope = .03, ns). This interaction was unexpected. Apparently, the less protesters identify with the others involved, the more their motivational strength is influenced by ideology motives, whereas, again, the motivational strength of strong identifiers remains high, irrespective of the level of their ideology motives. We will return to this issue in the discussion. The results found for the value-oriented protest event reveal a slightly different pattern. Again, the motivational strength of weak identifiers is strongly influenced by ideology motives (simple slope = .69, p < .001). However, in contrast to the former two interactions, motivational strength of strong identifiers increases (simple slope = .37, p < .001) as their ideology THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 82 participation motive increases. Thus, motivational strength of weak identifiers is strongly influenced by the level of their ideology motives; yet, contrary to power-oriented protest, the motivational strength of strong identifiers is stronger when ideology motives are stronger. The findings of these moderation analyses are contrary to our expectations. We expected that the influence of instrumental and ideology motives on motivational strength would be stronger for strong rather than weak identifiers, contingent on social movement context. Our studies reveal a more complex pattern. First of all, the motivational strength of weak rather than strong identifiers is strongly influenced by increasing instrumental and ideology motives. Next, the motivational strength of strong identifiers in the context of power-oriented protest is high, irrespective of the level of instrumental or ideology motives, whereas the motivational strength of strong identifiers in the context of value-oriented protest increases when their ideology motives increase. Hence, contrary to power-oriented protest where the level of instrumental and ideology motives does not influence the motivational strength of strong identifiers, the level of ideology motives does influence motivational strength for strong identifiers in the context of valueoriented protest. These findings suggest that the impact of instrumental and ideology motives on motivational strength differs for weak and strong identifiers contingent upon social movement context. One important finding is that, in the context of power-oriented protest, the motivational strength of strong identifiers remains unaltered by the (high) level of instrumental and ideology motives. We assumed, however, that identification with an organization staging power-oriented protest would influence instrumental rather than ideology motives, whereas identification with an organization staging value-oriented protest would influence ideology rather than instrumental motives. So far, this matter remains untested. We therefore conducted eight regressions to test the effects of identification on instrumental and ideology motives for weak and strong identifiers in the context of power- and value-oriented protest. Table 4.3 provides an overview of these regressions. 83 CHAPTER 4 Table 4.3 Regressions between identification and instrumental/ideology motives for weak and strong identifiers with organizations that stage power- and value-oriented protest. Power-oriented Motive Value-oriented weak strong weak strong identification identification identification identification Instrumental = .18 = .25*** = -.17 = -.03 Ideology = -.06 = .08 = .04 = .34*** Note: *** = p < .001 In the context of power- and value-oriented protest, identity motives of weak identifiers are not related to instrumental or ideology motives. Identity motives of strong identifiers in the context of power-oriented protest are much less related to ideology motives ( = .08, ns) than to instrumental motives ( = .25, p < .001). In the context of value-oriented protest, the data reveal the opposite pattern: identity motives are negatively, though weakly, related to instrumental motives ( = -.03, ns) and strongly related to ideology motives ( = .34, p < .001). This suggests that identification with an organization that stages power-oriented protest influences instrumental rather than ideology motives, whereas identification with an organization staging value-oriented protest triggers ideology rather than instrumental motives. In conclusion, identification with organizations staging power-oriented protest reinforces instrumental rather than ideology motives, whereas identification with organizations staging value-oriented protest reinforces ideology rather than instrumental motives. Furthermore, identification moderates the relation between instrumental and ideology motives and motivational strength in the context of power-oriented protest, and between ideology motives and motivational strength in the context of value-oriented protest. In other words, the relation between instrumental and ideology motives and motivational strength differs at different levels of identification. More specifically, the stronger the instrumental or ideology motives of weak identifiers, the stronger the motivation to participate in power-oriented protest; and the stronger the ideology motives of weak identifiers in the context of value-oriented protest, the more they influence motivational strength. Thus, the motivational strength of weak identifiers is strongly influenced by the level of instrumental and ideology motives, and both levels of instrumental and ideology motives are unrelated to identity motives. The pattern of strong identifiers is less clear-cut: in the context of power-oriented protest, the motivational strength of strong identifiers is invariably high independent of level of THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 84 instrumental and ideology motives. However, instrumental motives are influenced by identification with the union, whereas ideology motives are unrelated to this identification. This suggests that, although ideology motives play a vital role in motivational strength, they are not group-based, whereas instrumental motives are. In the context of value-oriented protest, the motivation of strong identifiers is reinforced by stronger ideology motives. Moreover, identity motives are related to ideology rather than instrumental motives. This suggests that ideology motives are group-based. We will return to this in the discussion. 4.2.5 Mediating effects of instrumental and ideology motives. We hypothesized that instrumental rather than ideology motives would explain the relation between identity motives and group-based anger in power-oriented protest events, whereas ideology rather than instrumental motives would explain the relation between identity motives and group-based anger in value-oriented protest events. To test our ideas concerning the mediational role of group-relevant appraisals on identity and group-based anger motives, we followed Baron and Kenny’s (1986) three-step procedure. As indicated, (1) the predictor should be significantly related to the mediator (identity motives are related to instrumental motives, = .42, p < .001), (2) the predictors should be related to the outcome variables (identity motives are related to group-based anger, = .08, p < .10) and (3) the mediating variable should be related to the outcome variable with the predictor included in the equation (instrumental motives are related to group-based anger controlling for identity motives, overview of the regressions for power-oriented protest. = .29, p < .001). Table 4.4 provides an 85 CHAPTER 4 Table 4.4 Regressions to test mediation of instrumental on identity and group-based anger motives in poweroriented protest: (1) identity on instrumental motives, (2) identity on group-based anger motives, and (3) group-based anger on instrumental controlled for identity motives. Power-oriented Instrumental Anger Anger Equation 1 Identity .42*** Equation 2 .08º Identity Equation 3 .28*** Instrumental -.01 Identity Model F 38.89*** 18.07*** 9.07*** df (1; 214) (1; 217) (2; 211) R2 .15 .08 .08 Note: º p < .10, ***p < .001 Thus, for the mediational model of the power-oriented protest the three mediation conditions were satisfied (cf. Baron & Kenny, 1986; Judd & Kenny, 1981). Most important, when identity and instrumental motives were included as predictors of group-based anger motives, instrumental motives predicted group-based anger significantly, whereas identity was reduced from = .08 to = -.01 (p =. 84), indicating that evidence for full mediation exists because the relationship between identity and group-based anger motives is no longer significant after controlling for the effects of instrumental motives (Baron & Kenny, 1986). An investigation of the indirect effect of instrumental motives on group-based anger revealed that instrumental motives reduced the effect of identity motives on group-based anger significantly, as the Sobel-Goodman test shows (Sobel z-value = 3.35, p < .001, direct: -.01, indirect: .08). Thus, the relation between identity and groupbased anger in power-oriented protest can be completely explained by instrumental motives. We assumed, however, that in power-oriented protest the relation between identity and group-based anger would be explained by instrumental rather than ideology motives. Following THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 86 the same strategy, we tested the mediational model with ideology as mediator. This analysis revealed no mediational effect of ideology between identity and group-based anger in poweroriented protest (Sobel z-value = .82, p = .41), indicating that protesters identifying with organizations staging power-oriented protests are angry for instrumental rather than ideological reasons. The same strategy was followed to test the mediational effects of ideology rather than instrumental motives on identity and group-based anger in value-oriented protest. In this mediation model, stronger identity motives are accompanied by stronger ideology motives ( = .26, p < .001); stronger identity motives are accompanied by stronger group-based anger motives ( = .17, p = .03); and ideology motives are related to group-based anger controlling for identity motives ( = .41, p < .001). Table 4.5 gives an overview of the regressions. Table 4.5 Regressions to test mediation of ideology on identity and group-based anger motives in value-oriented protest: (1) identity on ideology motives, (2) identity on group-based anger motives, and (3) group-based anger on ideology controlled for identity motives. Value-oriented Ideology Anger Anger Equation 1 Identity .26*** Equation 2 .17* Identity Equation 3 Ideology .41*** Identity .06 Model F 13.20*** 41.26*** 19.22*** df (1; 183) (1; 201) (2; 179) R2 .07 .17 .18 Note: *p < .05, ***p < .001 It is of primary interest, however, that, when ideology and identity were included as predictors of group-based anger, ideology predicted group-based anger significantly, = .41, p < .001, 87 CHAPTER 4 whereas the beta weight of identity motives was reduced from = .17 to = .06, p = .38. An investigation of the indirect effect of ideological motives on group-based anger revealed that ideological motives reduced the effect of identity motives on group-based anger significantly as the Sobel–Goodman test shows (Sobel z-value = 3.13, p < .001, direct: .06, indirect: .10). Thus, the path between identity and group-based anger was reduced significantly when ideology was included in the model, and the direct relation between identity and group-based anger became non-significant in the analysis. This suggests full mediation. What about the mediational influence of instrumental motives on the relation between identity and group-based anger in value-oriented protest? This analysis revealed no such mediational effect (Sobel z-value = -.79, p = .43). This suggests that the more people identify with organizations staging value-oriented protest, the more their group-based anger can be explained by ideological rather than instrumental reasons. See Figure 4.2 for a graphic overview of the observed mediational models of the power- and value-oriented protest events. power-oriented protest .08º Identity (-.01) .42*** Instrumental value-oriented protest Group-based anger .28*** (.29***) .17* Identity (.06) Group-based anger .26*** Ideology .41*** (.41***) Figure 4.2. Mediational model for power-oriented protest event and value-oriented protest event. All in all, the analyses show that the influence of instrumental and ideology motives on motivational strength is contingent not only upon identity, but also upon social movement context. Moreover, social movement context explains the origin of protesters’ anger. 4.2.6 Structural equation modelling. Structural equation modelling has one advantage over the previous analyses. It allows a direct test of the hypothesis that instrumental and ideology motives mediate between identity and group-based anger motives; and, although it is more difficult to test moderator models with structural equation modelling, we were able to incorporate the identification as moderator effect THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 88 and the social movement context moderator effect by using a series of multiple-group comparisons. In order to examine the models that represent our argument, we employed Analysis of Moment Structures (AMOS) software (Arbuckle, 1997). In building our proposed baseline model, we followed the development of social psychological research in protest participation. First, we tested a model comprising the instrumental and identity paths to protest participation (Simon et al., 1998), followed by a model encompassing the instrumental and group-based anger paths (van Zomeren et al., 2004). Next, we tested a model encompassing instrumental, identity and group-based anger motives, which we subsequently extended by adding an ideology pathway. The result was a model combining the instrumental, identity, ideology and group-based anger pathways to protest participation. The last step in building our proposed baseline model accounting for the strength of motivation to participate in protest was to include the mediating effects of instrumental and ideology motives on identity and group-based anger motives. Finally, two multiple-group comparison analyses were conducted to test the moderating effects of identification and social movement context. The multiple-group analysis of the moderating effects of identity was set up to test whether identity motives influence (direct or indirect) the motivational strength of strong identifiers whereas they do not for weak identifiers. The multiple-group analysis testing the moderating effects of social movement context was set up to test whether instrumental motives mediate between identity and group-based anger in the context of power-oriented protest and ideology motives on identity and group-based anger in the context of value-oriented protest. In multiple-group comparison, the baseline model is fitted separately to the covariance matrices of each social movement context (see Byrne, 2004). Taking the moderating effects of identity into account we hypothesize that all the paths from identity motives will be of minor importance for weak identifiers and thus insignificant, whereas they play an important role for high identifiers. Furthermore, taking the moderating effects of social movement context into account, we hypothesize that the ideological path in power-oriented protest and the instrumental path in value-oriented protest will be of minor importance in explaining the variance in motivational strength and thus insignificant. Consequently, applying an inductive approach, the insignificant paths will be fixed to zero. Yet, we hypothesize that fitting the baseline model to the weak and strong identifiers sample and the power- and value-oriented social movement context will reveal different insignificant paths, consequently the modified models will be different. If a constrained model fits the data as well as an unconstrained model, then the two models have comparable fit to the data and for parsimony reasons, that is, if the same amount of variation in 89 CHAPTER 4 motivational strength is explained with fewer parameters, the constrained model is preferred. We will elaborate on the method in the section on multiple-group comparison. SEM: Instrumental and identity pathways to protest participation. As indicated, Simon and colleagues (Simon et al., 1998; Stürmer & Simon, 2004; Stürmer et al., 2003) showed that protesters could take two routes to protest participation: an instrumental and an identity route. The instrumental route is taken when people participate because they expect that, by participation, they might solve a problem at affordable costs, and the identity path is taken when people identify with the others involved. According to Simon and colleagues, the two pathways are independent pathways to participation. The first model tested represents the direct effects of instrumentality and identity on motivation (see Figure 4.3). Note that in order to build a baseline model of protest participation, the alliance and the union protesters are taken together (N = 449). Multiple-group comparison analyses, however, are carried out for the weak and strong identifiers and the power- and value-oriented protest separately. instrumental .12** motivational strength identification .25*** Note. * = p < .05, * * = p < .01, *** = p < .001. Figure 4.3. Structural equation model, instrumental and identification pathway predicting motivation to participate in protest events. This model does not fit the data very well, with a significant chi-square value, 2 (1; 449) = 13.64, p < .001. Moreover, other fit indices also indicate poor fit: comparative fit index (CFI) = .76, Normed Fit Index (NFI) = .77 and root-mean square error of approximation (RMSEA) = .15 (see Hu & Bentler, 1999). In other words, although the pathways are significantly related to protest participation (instrumental = .12, p = .01 and identity = .25, p < .001), the overall model is not significant. Interestingly, the beta weight of the identity path is of the same THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 90 magnitude as the beta weights reported by Simon and colleagues (Simon et al., 1998; Stürmer & Simon, 2004; Stürmer et al., 2003) ranging from .22 to .26 (with an exception in the USA gay movement of .54). These beta weights represent the influence of identity on intention to participate in political protest. In one study they measure actual participation, resulting in a beta weight of .18 (Stürmer & Simon, 2004). However, the beta weight of the instrumental motive is rather low compared with the beta weights reported by Simon and colleagues. They report beta weights for instrumental motives between .08 and .20 for the collective motive, .10 and .38 for the normative motive, and .08 and .47 for the reward motive, whereas the beta weight of instrumental motives in our total sample is .12, .19 for the power-oriented and .10 for the value-oriented. We will return to this in the general discussion. Furthermore, the Squared Multiple Correlation (comparable with R2 in regression) of motivational strength that is used to judge the extent to which each of the independent variables contributes toward predicting the dependent variables is in this model .08; this is comparable with the explained variance in the study measuring actual participation in political protest (7%, Stürmer & Simon, 2004) Thus, both pathways uniquely predict protest participation, and the beta weights are comparable with the findings of Simon, Stürmer and colleagues. Moreover, the explained variance of motivational strength and actual participation is of the same magnitude. Thus, in general, the findings of Simon and Stürmer and colleagues are replicated.7 SEM: Instrumental and group-based anger pathways to protest participation. Van Zomeren et al. (2004) propose a different dual path model for protest participation. The routes to protest participation suggested by these scholars consist of an instrumental pathway and a group-based anger pathway. The instrumental pathway in this model is also taken by people who want to solve a problem or, as van Zomeren et al. (2004) argue, people engaged in problemfocused coping. The group-based anger pathway, however, is aimed at emotion-focused coping: protest participation is aimed at reducing the emotions attached to the problem. Van Zomeren and colleagues hold that these two pathways are two independent pathways to protest participation. Accordingly, our second model is a replication of the studies of van Zomeren et al. (2004) and represents the direct effects of instrumental and group-based anger on motivation (see Figure 4.4). 91 CHAPTER 4 instrumental .10* motivational strength .34*** anger Note. * = p < .05, * * = p < .01, *** = p < .001. Figure 4.4. SEM: instrumental and group-based anger path predicting motivational strength. This model does not fit the data very well either ( 2 (1; 449) = 16.32, p < .001, while the other fit indices indicated poor fit as well: CFI = .80, NFI = .80, and RMSEA = .19). The findings of van Zomeren et al. are replicated, in so far as the two pathways are significantly related to protest participation (instrumental = .10, p = .03 and group-based anger = .34, p < .001), but the overall model is not significant. The Squared Multiple Correlation (R2) of motivation is .13. Hence, the explained variance of instrumental and group-based anger on motivation is higher (.13) than the explained variance of instrumental and identity pathway (.08). Thus, the separate pathways of the two dual pathway models explain motivation to participate in protest events even though the models do not fit the data very well. SEM: Instrumental, identity, and group-based anger pathways to protest participation. Because all three routes to participation, instrumental, identity and group-based anger, appeared to be independent pathways to protest participation, we hypothesized that the three routes together would do better than each alone or any combination of two. This model does not fit the data very well either ( 2 (3; 449) = 33.62, p < .001, CFI = .73, NFI = .73, and RMSEA = .15). And, again, when group-based anger is introduced, the influence of instrumental motives on motivation to participate loses significance. Thus, two different statistical analyses (hierarchical regression reported in section 4.2.3 and structural equation modelling) reveal that group-based anger mediates between instrumental motives and 7 The more complex the model, the more likely a good fit (Garson, retrieved 19 July 2005 http://www2.chass.ncsu.edu/garson/pa765/structur.htm). Thus, probably due to a small number of variables the THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 92 motivational strength. The identity and group-based anger paths, however, are significantly related to protest participation ( = .10, p = .03 and = .34, p < .001, respectively) but the overall model is not significant (see Figure 4.5). instrumental .06 identity .22*** motivational strength .32*** anger Note. * = p < .05, * * = p < .01, *** = p < .001. Figure 4.5. SEM: instrumental, identity, and group-based anger path predicting motivational strength. The Squared Multiple Correlation of motivation is .16. This suggests that the explained variance of the models increases if the two dual path models are combined. Thus, although the models do not fit the data very well and despite the fact that the instrumental path is not significant, a combination of the three paths increases the explained variance. SEM: Instrumental, identity, group-based anger, and ideology paths to protest participation. First we tested a model with the four participation motives as independent paths to protest participation. Again, the data reveal a model with poor fit ( 2 (6; 449) = 107.09, p < .001, while the other fit indices indicate poor fit as well: CFI = .56, NFI = .56, and RMSEA = .19). By introducing the ideology path, the direct influence of the instrumental path on motivation remains insignificant. Adding ideology, however, does not change the impact of the other paths dramatically. Identity, ideology and group-based anger influence independently the motivation to participate in protest (see Figure 4.6). model does not reach levels of significance. 93 CHAPTER 4 instrumental identity .04 .20*** motivational strength ideology .41*** anger Note: .22*** *** = p < .001 Figure 4.6. SEM: instrumental, identity, ideology, and group-based anger path predicting motivational strength. The Squared Multiple Correlation of motivation is .19. Thus, including the ideology path again increases the variance explained by the model. Although, a model accounting for protest participation by simply accounting for direct effects does not fit the data, what will the model look like with the hypothesized indirect moderating and mediating effects? SEM: Baseline model accounting for motivation to participate in political protest. The model wherein we included the direct effects of the four paths to protest participation reveals that identity, ideology and group-based anger influence motivation to participate directly, whereas instrumental motives are completely mediated by group-based anger. To recall, we hypothesized that the relation between identity and group-based anger would be mediated by group-relevant motives; therefore we assume that ideology and instrumental motives function as mediators between identity and group-based anger. Figure 4.7 visualizes these pathways. This model fits the data very well, with an insignificant chi-square value, 2 (1; 449) = 1.72, p = .42. Moreover, other fit indices also indicate good fit: CFI = 1.00, NFI = .98 and RMSEA = .00. In other words, our hypothesized variance-covariance matrix does not differ significantly from the original matrix. Furthermore, the Squared Multiple Correlation (R2) of motivation in this model is .26. Thus, introducing the hypothesized mediating and moderating effects improved the model significantly. THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES identity 94 .20*** .17*** instrumental .15*** .14*** anger .22*** motivational strength .13*** .35*** ideology .31*** Note. *** = p < .001. Figure 4.7. SEM: baseline model predicting motivational strength. All the proposed paths are highly significant; this indicates that identity, group-based anger and ideology directly influence motivational strength. Furthermore, identity enhances both instrumental and ideology motives, which in their turn strengthen group-based anger motives. At two points the model departs from the one we proposed in Chapter 1. The data reveal a path between ideology and instrumental motives that we did not hypothesize. This suggests not only that people want to express their view but that they want to solve the problem as well. Next, we hypothesized a direct path from instrumental motives to motivation; the data reveal, however, that this relation is completely mediated by group-based anger. Apparently, the idea that participation might solve a problem at affordable costs alone is not sufficient; one needs to be angry as well. These two empirical adjustments seem to indicate that motives to participate are neither completely instrumental nor completely ideological. Multiple-group analysis: the impact of group identification strength. The above tested model does not take the moderating effects of group identification and social movement context into account. As indicated, we hypothesized moderating effects for group identification and for social movement context. First we will present the structural equation models for the moderating effects of group identification, followed by the moderating effects of social movement context. We hypothesized that group identification would reinforce motivational strength directly and indirectly via instrumental and ideology motives: among strong identifiers identity will 95 CHAPTER 4 impact directly on motivational strength, an indirectly because instrumental and ideology motives may mediate between identity motives and group-based anger. Among weak identifiers we hypothesized no such direct or indirect effects of identity, but, we assumed that the motivational strength of weak identifiers would be contingent on instrumental and ideology concerns. To test these hypotheses the baseline model will be tested for two samples created by median split of identity motives separately8. Next, the insignificant paths will be forced to zero and the modified models will be tested again for the two samples separately. The third and last step is to compare the chi-squares of the original baseline model with those of the modified models. We hypothesized that the model for weak identifiers fits the data significantly better if the paths from identity were fixated to zero while the model for strong identifiers is equal to the baseline model. The key statistic is the change in chi-square from the baseline submitted to the separate samples and the modified models. If the chi-square difference is not significant, then the baseline and modified model have comparable fit to the data and for parsimony reasons, the modified model will be preferred. We start of with a test of the baseline model submitted to the sample of the weak identifiers (see Figure 4.8, to the left) and the strong identifiers (Figure 4.8 to the right). identity identity -.01 .28* instrumen .11* -.01 tal anger .34*** .32* .35* instrumen .16** motivation tal anger .09* .45*** .14* .43*** .54*** ideology Weak identifiers .12* .30*** .19*** ideology Strong identifiers Note. * = p < .05, * * = p < .01, *** = p < .001. (dashed lines represent insignificant paths) Figure 4.8. Baseline model submitted to weak and strong identifiers. 8 Using a median split of the union sample (m = 4.46) and the alliance sample (m = 3.76), each participant was categorized as either a weak identifier or strong identifier. motivation THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES The baseline model for the weak identifiers fits the data well ( 2 96 (2; 148) = 4.45, p = .49, CFI = 1.00, NFI = 1.00 and RMSEA = .00. As hypothesized, the paths from identity to ideology ( = .01, ns) and from identity to instrumentality ( = -.01, ns) are not significant. However, contrary to our hypothesis, identification with the social movement organizations does—even for weak identifiers—reinforce motivational strength. Testing the baseline model for strong identifiers shows acceptable goodness of fit indices, all significant paths, but a significant chi-square. This implies, contrary to our hypothesis, a rejection of the baseline model submitted to the strong identifiers sample ( 2 (2; 271) = 5.96, p = .05, CFI = .96, NFI = .95 and RMSEA = .08). The next step in the multi group analyses was to force the statistically insignificant paths for the weak identifiers samples to zero (see Figure 4.9). Forcing the identity/ideology path and the identity/instrumental path to zero in the weak identifiers sample revealed a modified model with very good fit as well ( 2 (4; 148) = 3.15, p = .53, CFI = 1.00, NFI = .97 and RMSEA = .00). identity identity 0 .28* instrumen .11 tal 0 anger .34*** .35* instrumen .14** motivation .32* .33** tal anger .09* motivation .45*** .14* .43*** .54*** .12* ideology .28*** .19*** ideology Weak identifiers Strong identifiers Note. * = p < .05, * * = p < .01, *** = p < .001. Figure 4.9. Modified models submitted to weak and strong identifiers. To test whether the baseline model or the modified model fits the sample of the weak identifiers better than the baseline model, the change in chi-square from the baseline model to the modified model is calculated: 2 ( df. =2; 148) = 1.30, p > .50. This chi-square difference is not CHAPTER 4 97 significant, indicating that the baseline model and the modified model fit the data equally well. However, the more parsimonious model is to be preferred. This suggests that motivational strength for weak identifiers is influenced directly by identification but not by the indirect effects of identification on instrumentality and ideology motives. Among the strong identifiers sample the baseline model fitted the data adequately. The various paths were significant but the chi-square was significant as well. Therefore the model was rejected. In an attempt to improve the model we added a direct path from identity motives to group-based anger, as suggested by the modification indices (see Figure 4.9). The path coefficient is = .33 (p = .02) indicating that the more people identify with social movement organizations the angrier they are. This implies a replication of the findings of Yzerbyt (2003) and colleagues that the more the group is in me (i.e., the stronger the group identification) the more I feel for us. The resulting model submitted to the sample of the strong identifiers revealed a good fit ( 2 (1; 271) = .67, p = .41, CFI = 1.00, NFI = .99 and RMSEA = .00). This implies, that for people with strong ties to a social movement organization, identity motives reinforce instrumental and ideological motives, which in turn make them angrier. Moreover, identity motives create a shortcut to anger: people who strongly identify with a social movement organization do not per se need instrumental or ideological motives to become angry. At face value, this modified model seems to be an improvement, and the chi-square difference confirms this ( 2 ( df. =1; 271) = 5.29, p < .05). Note, this change is significant, indicating that adding a direct path from identity to group-based anger improves the model significantly. At two points the model departs from the hypotheses on the moderating effects of identification we proposed in Chapter 1. The data reveals that—even for weak identifies— identity motives impact directly on motivational strength, and, strong identification with social movement organizations creates a shortcut to anger. These models reveal clearly that weak and strong identification with a social movement organization impact on the motivational pattern that spur people to take the street. Neither emotions nor motives of the weak identifiers are influenced by identification with the social movement organization; therefore the motivation to become involved in the struggle for weak identifiers can be seen as individualistic strategies to protest personal interests (or more opportunistic strategies, Ellemers, Spears, & Doosje, 1999). The emotions and motives for strong identifiers, on the other hand, seem to be influenced by their—strong—ties to the social movement organization, suggesting that a threat to collective group interests result in shared grievances and emotions which reinforces the motivation to support the group (or more solidarity THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 98 strategies, Ellemers et al., 1999). We will return to the various effects of identification on protest participation in the general discussion. Multiple-group analysis: the impact of social movement context. Thus far, we have not taken the difference in mobilizing context into account, but we assumed that the motivational dynamics might differ depending on the action orientation of the protest. As indicated, we hypothesized that instrumental motives would prevail in the context of power-oriented protest and ideology in the context of value-oriented protest, whereas identity and group-based anger motives would be equally important for both action orientations. This points to an identity path via instrumentality to group-based anger for power-oriented protest, and an identity path via ideology to group-based anger for value-oriented protest. To test these hypotheses, the above-presented baseline model (that is, assessed both on the power- and valueoriented protest sample) will be tested for the two samples separately. The second—inductive— step is to force the insignificant paths to zero and test the modified models for the two samples. The third and last step is to compare the chi-squares of the original baseline model tested for the separate samples with the modified models for the separate samples. In other words, does the baseline model significantly improve if it is adapted for different social movement contexts? Hence, does the model of the power-oriented protest event significantly improve if the ideology path is fixated to zero and does the model of the value-oriented protest event significantly improve if the instrumental path is fixated to zero? The key statistic is the change in chi-square for the different models. If the chi-square difference is not significant, then the two models have comparable fit to the data and, for parsimony reasons, the modified model is preferred. We start off by testing the baseline model for the power-oriented protest. Testing the baseline model for the power-oriented protest shows excellent fit ( 2 (2; 449) = .33, p = .85, CFI = 1.00, NFI = 1.00 and RMSEA = .00, Squared Multiple Correlation [R2] of motivation = .18). As hypothesized, the path from identity to ideology is not significant ( = .06, p = .30), whereas the path from identity to instrumental motives is ( = .47, p < .001), indicating that identification with social movement organizations staging power-oriented protest shapes instrumental rather than ideological participation motives. Contrary to our expectations, these structural equation analyses (like the hierarchical regression reported in section 4.2.3) reveal that in this power-oriented protest event protesters take the ideology path as well. We hypothesized that in power-oriented protest the instrumental rather than the ideology path would prevail. Apparently, expressing one’s view and venting one’s anger is more important in the context of power-oriented protest than we assumed. We will return 99 CHAPTER 4 to this in the discussion. The path from ideology to group-based anger is mediated by instrumental motives, indicating that the path from ideology to motivation via group-based anger is partly explained by instrumental motives. Forcing the identity/ideology path to 0 revealed a modified model with excellent fit as well ( 2 (3; 449) = 1.21, p = .75, CFI = 1.00, NFI = .99 and RMSEA = .00). See Figure 4.10. identity .19*** .47*** instrumental 0 .18*** .20*** anger .30*** .13** motivational strength .21*** ideology Note: * * = p < .01, *** = p < .001 Figure 4.10. Baseline model modified for power-oriented protest. Then, what model fits the data of the power-oriented protest event better, the baseline model or the modified model with an identity/instrumental path and an ideology component? As indicated, the key statistic of interest here is the change in chi-square of the baseline model submitted to power-oriented protest sample compared with the modified model: 2 ( df. =1; 213) = .88, p > .70. This chi-square difference is not significant, indicating that the baseline model and the modified model have comparable fit to the data. However, for parsimony reasons the modified model is preferred. Hence, the model with an instrumental path and an ideology component is more parsimonious and fits the data equally well. Moreover, the finding of the mediational analysis in section 4.2.5, that instrumental rather than ideology motives mediate the relation between identity and group-based anger in the context of power-oriented protest, is replicated with structural equation analysis, indicating that the group-based anger of protesters in a power-oriented protest is influenced by instrumental THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 100 motives. These results suggest that instrumental rather than ideology motives function as group-relevant motives in power-oriented protest. A final remark about the modified model in the context of power-oriented protest is about the ideology path. As indicated, we did not hypothesize an ideology path in power-oriented protest. The data reveal, however, a direct path from ideology to motivational strength, and an indirect path via group-based anger (people who want to express their view and vent their anger) and an indirect path via instrumental motives to group-based anger, indicating that, in the context of power-oriented protest, the group-based anger of protesters taking the ideology path may be influenced by instrumental motives. Subsequently, the baseline model was tested for the value-oriented protest sample ( 2 (2; 449) = 2.31, p = .32, CFI = 1.00, NFI = .98 and RMSEA = .03, Squared Multiple Correlation of motivation = .33). As hypothesized, three paths are non-significant: identity to instrumental motives ( = -.09, p = .18), ideology to instrumental motives ( = 10, p = .15), and instrumental to group-based anger motives ( = .05, p = .48), indicating that an identity/ideology path to value-oriented protest prevails. Forcing these instrumental paths to zero creates a modified model with an identity/ideology pathway to value-oriented protest. This modified model fits the data well ( 2 (5; 449) = .18, p = .32, CFI = .98, NFI = .95 and RMSEA = .05, and the Squared Multiple Correlation of motivation = .34). Figure 4.11 provides an overview of this modified model. identity 0 .23*** instrumental .0 .14* 0 anger .42*** .30*** motivational strength .44*** ideology Note. * = p < .05, *** = p < .001. Figure 4.11. Baseline model modified for value-oriented protest. Does this modified model fit the data of the value-oriented protest sample better than the baseline model? The change in chi-square is 5.30, whereas the change in degrees of freedom is 3, which is CHAPTER 4 101 not significant (p > .20), indicating that the baseline model and the modified model have comparable fit to the data. However, for parsimony reasons the modified model is preferred. Thus, a model with only an ideology path to protest participation accounts as well as the complete baseline model for the variance in strength of motivation. Identification with an organization staging value-oriented protest is related to ideology rather than instrumental participation motives; this indicates that ideology motives are grouprelevant for organizations staging value-oriented protest. Moreover, ideology rather than instrumental motives mediate between identity and group-based anger. This suggests that identification with an organization staging value-oriented protest shapes ideology rather than instrumental motives, which in turn strengthens group-based anger. The mediational analysis of section 4.2.5 is replicated in the structural equation analysis. Again, the data reveal that identity processes influence not only the fact that group members are angry, but also what group members are angry about. Hence, in the context of power-oriented protest, instrumental motives mediate between identity and group-based anger, whereas in the context of value-oriented protest ideology motives mediate between identity and group-based anger; this indicates that the mediational effects of instrumental and ideology motives on identity and group-based anger are contingent upon social movement context. This result replicates and extends the findings of Yzerbyt et al. (2003), who showed that group-based anger mediates on group identification and motivational strength. 4.3. Conclusions The results of these studies provide insights into the motivational dynamics of protest participation. Instrumental, identity, ideology and group-based anger participation motives appear to influence the strength of motivation to participate in political protest and work together, as a function of the action orientation of the protest. The results take us considerably beyond the insights already available in the literature. Various pathways have been proposed and tested, but to our knowledge no single study has tested the various pathways together or attempted to assess the relative weight of various paths, or how group identification strength or contextual variation influences the relative weight. 4.3.1 Model accounting for the strength of motivation to participate in political protest. Previous research suggested the possibility of an identity path (Simon et al., 1998) and a groupbased anger path (van Zomeren et al., 2004) in addition to the instrumental path (Klandermans, THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 102 1984) to protest participation. Our studies replicated and extended these findings by integrating the two dual-path models into a single model and adding a fourth path, ideology. The findings show that it is fruitful to add a fourth pathway to the existing instrumental, identity, and group-based anger paths. In the context of value-oriented protest the ideology motive added 16 % to the variance already explained and even in the context of the power-oriented protest ideology motives added 5 % to the variance already explained. Indeed, people’s ideals and values generate passionate politics and spur people to participate in political protest. Integrating the four paths in a single model revealed an unexpected finding regarding the instrumental path: the influence of instrumental motives on the strength of motivation was completely mediated by group-based anger. Thus, whereas identity and ideology motives directly feed into the strength of motivation, instrumental motives need to be translated into anger in order to motivate people to take to the streets. One might wonder whether this is due to an integration of the two dual path models or resulting from this specific protest campaign. We will further elaborate on the role of instrumentality in the general discussion. 4.3.2 The influence of identification processes on protest participation. Group identification processes are known to influence protest participation directly, our study refines and enlarges the understanding of identification processes by hypothesizing and testing indirect effects of group identification processes. We replicate the finding that identity motives feed directly into the strength of motivation, indicating that identification with others involved influences the strength of motivation sufficiently to take to the streets; but identity motives also play a vital role in the indirect path taken to protest participation. Our hypothesis regarding the mediational role of instrumental and ideology motives, was confirmed. Mediational analyses revealed that the relation between identity and group-based anger in the context of power-oriented protest is completely mediated by instrumental motives, and in the context of value-oriented protest by ideology motives. Thus, identification not only evokes group-based anger but also shapes protesters’ participation motives. Identification with an organization staging power-oriented protest makes people angry for instrumental rather than ideology reasons, whereas identification with an organization staging value-oriented protest makes people angry for ideology rather than instrumental reasons. Our hypothesis regarding the moderating effects of identity was also confirmed. The routes to protest participation of weak identifiers starts of at instrumental and ideological motives, suggesting that identity motives do not reinforce instrumental and/or ideological motives or group-based anger. However, contrary to our expectations, identity motives do—even CHAPTER 4 103 for weak identifiers—directly feed into the strength of motivation. This is consonant with findings of Veenstra and Haslam (2000) who showed that in times of threat weak identifiers are more inclined to participate in collective action. In the motivational pattern of strong identifiers, on the other hand, identification with a social movement organization reinforced instrumental, ideological, and group-based anger motives. Moreover, contrary to our hypothesis, identity does not only influence group-based anger indirectly via instrumentality and ideology motives, but also directly. Thus, the stronger the ties of strong identifiers with a social movement organisation the angrier they are, which in turn reinforces their motivation to participate in political protest (cf Yzerbyt et al., 2003). The moderation models of identity reveal clearly that weak and strong ties with a social movement organization make a difference for the motivational patterns that spur people to take the street. Instrumental, ideological, and group-based anger motives of weak identifiers are uninfluenced by identification with the social movement organization. The motivational pattern of weak identifiers seems to stem from a salient personal identity, accordingly, weak identifiers experience grievances as a threat to their personal identity and they “feel for me”. This suggests that weak identifiers’ motivational strength to participate in collective action is a by-product of their desire to protect their own personal interests (Wright et al., 1990). The motivation for weak identifiers to become involved in the struggle results from individualistic, or more opportunistic strategies (Ellemers, Spears, & Doosje, 1999). The emotions and motives for strong identifiers, on the other hand, seem to be influenced by their—strong—ties to the social movement organization, suggesting that a threat to collective group interests results in shared grievances and emotions (“I feel for us”). For strong identifiers, group interests are internalized and adopted as self-interests (Turner et al., 1987). This suggests that strong identifiers’ motivational strength to take the streets is reinforced by the motivation to support the group. The motivation for strong identifiers to take part in collective action on behalf of the group results from more solidarity strategies (Ellemers et al., 1999). Several protest scholars stress the role of identity in spurring protest participation (Kelly & Breinlinger, 1996; Reicher, 1984; Simon et al., 1998, Stryker et al., 2000; Stürmer et al., 2003). Our studies reveal that protest participation is not only stemming from commitment and solidarity. It is also the values, interests, and emotions held about the social movement organization, which are more or less incorporated in the individual’s collective identity that spur protest participation. We see this as further evidence of the central role of identity in protest participation. We will return to the various effects of identification on protest participation in the general discussion. THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 104 4.3.3 The influence of social movement context on motivational patterns. Our research design enabled us to compare the motivational patterns of two protests organized by different social movement organizations. We applied the concept of action orientation proposed by Turner and Killian (1987). Based on Turner and Killian’s conceptualization, we defined the trade unions’ demonstration as power-oriented and the Turn the Tide demonstration as valueoriented. For power-oriented protests, we found two direct paths to protest participation: both identity and ideology motives are directly translated into strength of motivation. Thus, identifying with others involved or wanting to express one’s view in itself is sufficient to motivate people to participate in power-oriented protest. We found two indirect paths as well, one from identity via instrumentality to group-based anger; this implies that the more people identify with the union, the more they are spurred by instrumental participation motives, and this in turn reinforces group-based anger. The other indirect path was from ideology motives via instrumental motives to group-based anger; this indicates that the stronger someone’s ideology motives, the more he or she is pushed by instrumental motives, which makes him or her even angrier. Thus, in power-oriented protest, instrumental motives mediate the relation between both identity and ideology motives on the one hand, and group-based anger motives on the other. Whether a protester starts from ideology or identity motives seems to depend on the extent to which he or she identifies with the union. It turned out that the less protesters identify with others involved, the more ideology and instrumental participation motives play a role in their motivation to participate. Regarding motivation to participate in a value-oriented protest, we found two direct paths and one indirect. The two shortcuts to participation are the same as in power-oriented protests, both identity and ideology motives in themselves are sufficient to generate a strong motivation to participate in protest. The indirect path goes from identity via ideology motives to group-based anger motives. In other words, people who identify with an organization staging value-oriented protest are more likely to be spurred by ideology motives; this in turn makes them angrier. A comparison of the two motivational patterns reveals that the two direct paths to either power- or value-oriented protests are the same, both identity and ideology motives on their own are strong enough to create sufficient motivation to take to the street. This suggests that these motives are strong, yet different, motivators. The motivation of protesters taking the ideology path to protest stems from integrity maintenance and emotional catharsis (Klandermans, 2003; 2004), whereas an identity path is spurred by identification, solidarity with other group members CHAPTER 4 105 involved and a felt inner obligation to behave as a “good” group member (Simon et al., 1998; Stürmer et al., 2003). The motivation on the ideology path is about my integrity and my emotions, whereas the motivation on the identity path is about identification with others, solidarity with others and a felt inner obligation evoked by others. It appears that the motivation to take the ideology path stems from oneself, whereas the motivation to take the identity path stems from others. In describing types of motivations to enlist as AIDS volunteers, Omoto and Snyder (1993) find the same distinction and they describe these motivations as self-focused and other-focused motivations. We will return to this in the general discussion. Whereas the direct paths to both types of protest are the same, the indirect paths differ. We start with the indirect ideology path. The ideology path in the context of power-oriented protest is mediated by instrumental motives, whereas instrumental motives are unimportant on the ideology path to value-oriented protest. Thus, protesters employing an ideology motive in the value-oriented protest participate because they want to express their view and vent their anger, whereas protesters employing an ideology motive in power-oriented protest want to solve a problem as well. Hence, although both power- and value oriented protest events attract people taking an ideology pathway, the influence of ideology on participation is almost twice as strong in the value-oriented protest as in the power-oriented protest. Testing the baseline model in different social movement contexts offered the opportunity to study why people participate and what makes them angry. Moreover, it helped to show that identification with organizations staging either power-oriented or value-oriented protest events makes a difference for protesters’ motives and the reasons why they are angry. In the context of power-oriented protest, identity motives moderate the influence of instrumental participation motives on motivation to participate, in contrast to value-oriented protest where identity moderates the influence of ideology motives on motivation to participate in protest. Consequently, instrumental rather than ideology motives in power-oriented protest, and ideology rather than instrumental motives in value-oriented protest, enhance the influence of group-based anger on motivational strength. Indeed, it appears fruitful to study the influence of group identification on political participation in a comparative design as Reicher (2004) indicates, “what we need to explain, then, is not the inevitable occurrence of any particular form of intergroup relations irrespective of context but rather the flexibility of behavior [cognitions, motives, and feelings, JvS] across differing contexts” (p. 925). In this chapter we have discussed the impact of context by characterizing the two protests as power-oriented and value-oriented by testing hypotheses derived from this distinct orientation. THE RELATIVE WEIGHT OF PARTICIPATION MOTIVES 106 In the next chapter we will concentrate on the psychological mechanisms that help to explain why some individuals end up taking the one route to protest while others take the other. The mechanisms we proposed in Chapter 1 were regulatory fit and regulatory focus. We will translate the characterization in terms of action orientation we borrowed from social movement literature into a characterization in terms of regulatory focus. On the basis of that characterization we will test our hypotheses regarding regulatory fit and regulatory focus. 107 Chapter 5. Regulatory focus and political protest participation. In the previous chapter we demonstrated that the motivational pattern of protesters varies if the action orientation of the protest event varies: in power-oriented protest, the instrumental path was found, combined with either identity or ideology motives, whereas in value-oriented protest the ideology path (whether in combination with identity motives or not) prevailed. The question remains as to why people attend one protest event rather than another. As stated, we conducted two surveys in two different squares at protests organized by two different movements at exactly the same time against the same budget cuts proposed by the same government. This Most Similar Systems Design is interesting from a social psychological point of view, because each cluster of organizations emphasized differing aspects of the proposed government policies. Indeed, their interpretation of “what’s going on?”, “who is to be blamed?” and “how we are going to solve it?”, in other words, their collective action frames, differed. Why would one collective action frame resonate more than another for someone? In this chapter we will try to answer this question employing the concept of regulatory fit, that is, the notion that information that is consistent with an individual’s regulatory focus has more impact than information that is inconsistent (Higgins, 1997; 1998). We will investigate whether a collective action frame constructed in promotion terms is more persuasive for promotion-focused people, while a collective action frame in prevention terms is more persuasive for prevention-focused people. Indeed, although their social and political environment provides an abundance of frames, regulatory fit may be one of the explanations as to why people adopt certain frames while neglecting others. Assessment of a regulatory fit between collective action frames of social movement organizations and individual self-regulation is an important first step in the explanation of political participation as a function of self-regulatory processes. Yet, this first step does not show us whether regulatory focus moderates the individual motivational mechanisms that lead to protest participation. Indeed, we still do not know whether regulatory focus influences what participation motive will be employed. In the second half of this chapter we will elaborate on this latter question. First, we elaborate on the regulatory fit between the triad of individual protesters, organizations and their collective action frames, and then we discuss regulatory focus as the steering mechanism for individual motivational dynamics explaining protest participation. REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 108 5.1 Regulatory fit: social movement context and regulatory focus. Why, in the first place, are people attracted to certain social movement organizations rather than others? Why, in the second place, are social movement organizations in times of intergroup conflict inclined to emphasize various aspects and interpretations in their diagnosis and prognosis? And why, finally, are people inclined to participate in one protest event rather than another? We argue that the attractivity of social movement organizations is not random. In a similar vein, collective action frames are not randomly defined, nor do protesters appear at protest events randomly. We will describe the dynamic context of recruitment, persuasive collective action frames and motivation to participate in protest staged by social movement organizations as an instance of regulatory fit. For argument’s sake, we will start off with recruitment and polarization due to intergroup conflict. Yet, our data only allow us to test the regulatory fit between collective action frames and regulatory focus of individual protesters. 5.1.1 Group attractivity and regulatory focus. Why are people attracted to one social movement organization rather than another? Indeed, what shapes people’s choice of one ingroup rather than another? Klandermans (1997) argues that, in a recruitment context, the same two steps should be taken as with every form of participation: “people must be targeted first and then motivated” (p. 67). Recruitment on the part of social movement organizations does not take place in isolation but by people interacting with other people in informal circles, primary groups and friendship networks. These informal structures of everyday life play an important role in movement mobilization. Kinship and friendship networks have been shown to be central to understanding movement recruitment (Klandermans & Oegema, 1987; Snow, Zurcher, & Ekland-Olson, 1980). In Indeed, being asked to become a member of a group is the first step, but what accounts for the motivation to actually become a member? In other words, what makes this group an attractive group? The group attractivity of social movement organizations is (at least partly) determined by what the organizations want to accomplish, in other words, their mission, goal or ideology. Missions, goals and ideologies may be stated in a more prevention- or a more promotion-oriented manner. In fact, Shah et al. (2004, p. 444) hypothesize that: “the needs to attain achievement and security at the individual or collective level may affect ingroup choice by defining the characteristics one looks for in groups as well as the characteristics one seeks to avoid. Prevention-oriented individuals, for instance, CHAPTER 5 109 may avoid ‘dangerous’ groups and may seek out ‘safe’ groups that may better provide security at both the individual and group level. Alternatively, promotionfocused individuals may be attracted to ‘upwardly mobile’ groups that show promise to grow as a whole, while offering possibilities for personal advancement, and avoid groups that offer little possibility for growth” Therefore we hypothesize that prevention-focused individuals are attracted by social movement organizations with missions, goals and ideologies stated in prevention rather than promotion terms, and promotion-focused individuals are attracted by social movement organizations with missions, goals and ideologies stated in promotion rather than prevention terms. Once people are recruited, the effects of ingroup identification are pervasive and powerful, even more so for acquired identities (i.e., identities adopted by choice, in contrast to ascribed identities which are quite difficult to change) because group identification tends to increase in strength when group membership is voluntary (Huddy, 2003). As social movement identities are typical examples of acquired identities, we hold that identification processes with social movement organizations are pervasive and powerful. The findings in classic and recent research indicate that people who identify with a group (1) perceive themselves to be more similar to each other (Allen & Wilder, 1975; 1979; Mackie, 1986); (2) are more likely to act cooperatively (Back, 1951); (3) feel a stronger need to agree with group opinion (Deutsch & Gerard, 1955; Mackie, Gastardo-Conaco & Skelly, 1992; Wilder, 1990); (4) perceive ingroup messages to be of higher quality (Brock, 1965; Mackie, Worth & Asuncion 1990); and (5) conform more in both behaviour and attitude (French & Raven, 1959; Wilder and Shapiro, 1984) even more strongly in time of intergroup conflict due to polarization of group attitudes and behaviour (Mackie & Cooper, 1984). 5.1.2 Collective action frames and regulatory focus. According to Rubin, Pruitt and Kim (1994), escalation of conflict occurs when a group is faced with aggression, or when one side perceives the other as the cause of loss or unfulfilled aspirations. As tensions rise and intergroup communication becomes channelled through more antagonistic lenses, “relevant ingroup norms [attitudes and behaviour, JvS] are likely to become more extreme so as to be more clearly differentiated from outgroup norms [attitudes and behaviour, JvS], and the within group polarization will be enhanced” (Brown, 1980, p. 154). Social action as a reaction to perceived losses or unfulfilled aspirations is the raison d’etre of social movement organizations. According to Klandermans (1984), social movements face two separate mobilization challenges: consensus mobilization, persuading people of the goodness of the cause, and action mobilization, actually bringing people onto the streets. Indeed, REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 110 in the struggle for the good cause, within-group polarization of relevant ingroup norms, attitudes and behaviour may play an even more important role for social movement organizations by their tendency to try to win bystanders to their side, forcing people to take sides and engage in mobilization. Relevant ingroup norms, attitudes and behaviours in terms of prevention and promotion may be, for example, that promotion-oriented groups are oriented towards advancement and accomplishment, whereas prevention-oriented groups are oriented toward security and responsibility. Moreover, promotion-oriented groups are concerned with the presence or absence of positive outcomes (gains and non-gains), whereas prevention-oriented group are concerned with the absence or presence of negative outcomes (non-losses and losses). Indeed, the abovestated reasons articulated by Rubin et al. (1994) for escalation of group conflict are a good example. Escalation of intergroup conflict occurs when one side perceives the other as the cause of a loss (i.e., prevention-oriented) or of unfulfilled aspirations (i.e., promotion-oriented). Because relevant ingroup norms, attitudes and behaviours will polarize in times of intergroup conflict, it seems reasonable to suggest that intergroup conflict intensifies the prevention characteristics of prevention-oriented social movement organizations and the promotion characteristics of promotion-oriented social movement organizations. Participating because of common interests requires a shared interpretation of who should act, why and how. Movements affect such interpretations by the information they disseminate. Social movement scholars use the concept of framing (Gamson, Fireman & Rytina, 1982; Snow, Rochford, Worden & Benford, 1986) to analyze this process of meaning construction. Motivation to act is grounded in accounts of why we should act—and often takes the form of a story about who we are, where we have been, where we are going, and why we want to get there. In Gamson’s (1992, p. 7) words, a collective action frame is “a set of action-oriented beliefs and meanings that inspire and legitimate social movement activities and campaigns”. According to Gamson, collective action frames consist of three components: (1) injustice, referring to moral indignation, a so-called hot cognition laden with emotions; (2) agency, or the consciousness that it is possible to alter conditions or policies through collective action; and (3) identity, referring to a “we” in opposition to some “they” who have different interests or values. We assume that defining a collective action frame is not a random process either. In contrast, we hold that prevention-oriented social movement organizations are inclined to define their collective action frame in prevention rather than promotion terms, whereas promotion-oriented social movement organizations are inclined to define their collective action frame in promotion rather than prevention terms. As previously mentioned, promotion-oriented frames are oriented towards CHAPTER 5 111 advancement and accomplishment, whereas prevention-oriented frames are oriented toward security and responsibility (Higgins, 1997). Promotion-oriented frames can also be defined in terms of the presence or absence of positive outcomes (gains and non-gains), whereas preventionoriented frames are defined in terms of the absence or presence of negative outcomes (non-losses and losses). 5.1.3 Frame alignment and regulatory focus. Individual members of a collectivity incorporate a smaller or larger proportion of the interpretations provided by “their” organizations; but there is an abundance of frames in our social and political environment, so why would people adopt certain frames while neglecting or paying less attention to others? Snow et al. (1986) propose frame alignment, a process of making individual and organizational frame fit. By frame alignment, Snow et al. mean the linkage of individual and social movement organization interpretive orientations such that: “some set of individual interests, values and beliefs and social movement organizations’ activities, goals and ideologies are congruent and complementary” (p. 464). In other words, individuals’ ideas line up with movement ideas. The concept of resonance is relevant to the issue of the effectiveness or mobilizing potency of frames, thereby attending to the question of why some framings seem to be effective or resonate, while others do not (Snow & Benford, 1988). A successful process of frame alignment results in a fit between the collective action frame of an organization and that of an individual, and this enhances the likelihood that this individual will participate in a protest event staged by this organization (Snow et al., 1986). The fit as a result of frame alignment is akin to the fit concept of regulatory focus theory (Higgins, 1997; 1998), hence a match between information and an individual’s regulatory state. Therefore, again, we hold that it is not accidental that people adopt certain collective action frames while neglecting others. We propose that collective action frames defined in prevention terms are more persuasive for prevention- than promotion-focused people, whereas collective action frames defined in promotion terms are more persuasive for promotion- than prevention-focused people. In summary, we propose a dynamic of mutual influence between individual regulatory focus and promotion- and prevention-oriented intra- and intergroup processes. People prefer to make decisions in a manner that fits their individual regulatory focus; we assume that this also applies to group-related decisions. Indeed, a fit between individual regulatory focus and regulatory group processes is preferred. People prefer promotion-oriented group processes when in a promotion-focus, and prevention-oriented group processes when in a prevention-focus. If group processes are shaped and influenced by regulatory processes, then regulatory processes REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 112 should play a role in social movement organizations too. Indeed, processes like recruitment, group identification, group polarization and definition of collective action frames may well be explained in terms of regulatory focus. Our data enable us to assess the fit between the regulatory characteristics of a collective action frame and the regulatory focus of individual protesters; in other words, a regulatory social movement organization-participant fit; that is to say, a fit of the regulatory focus of the collective action frame and the regulatory focus of the individual protester. A promotion-oriented collective action frame fits a promotion-focused individual protester. This explains why we expect promotion-focused individuals to participate in promotion-oriented protest events, whereas we expect the prevention-focused individual to take part in prevention-oriented protest. 5.1.4 Determining regulatory characteristics of the collective action frames. Due to the combined working of recruitment, network, group identification, polarization and regulatory fit, protesters end up protesting at one protest event rather than another. As indicated, our data do not allow us to test the effects of recruitment, networks and polarization, but offer the possibility to test our ideas about regulatory fit. Is it indeed possible to characterize campaigns as more or less promotion- or preventionoriented? The general budget cuts of the government were comprehensive, and the two movements emphasized different aspects of the proposed policies. Their diagnosis did not differ much; both organizations described the budget cuts as an erosion of the Dutch social security system. However, for the trade union federations, the most controversial proposed policy changes were: “the ‘pre-pension’ scheme [early retirement, JvS] and social security” (www.FNV.nl), whereas Turn the Tide was “against the current harsh rightist climate in The Netherlands and in particular against the asocial government policies”9. The prognosis of both organizations seemed to differ. The union prognosis agreed that changes in the social system were necessary, but in a different manner (D. Terpstra, 2 October, 2004, Amsterdam: “we are able too, to sway public opinion, together with the government. We too, including the people here at the Museumplein, know that changes are inevitable. Yet, in a different manner than proposed by the government)10. In other words, the union prognosis is about preventing social change in the way proposed by the government. We speculate that the 9 Original text: tegen het huidige harde rechtse klimaat in Nederland en in het bijzonder tegen het asociale kabinetsbeleid ”, retrieved from: www.keerhettij.nl, 1 October, 2004. 10 Original text: “wij kunnen ook, samen met de regering zorgen voor draagvlak. Ook wij, ook die mensen hier op het Museumplein weten dat er veranderingen noodzakelijk zijn. Maar dan op een andere manier dan het kabinet voorstelt”, retrieved from: www.nederlandverdientbeter.nl. CHAPTER 5 113 collective action frame of the trade union federations is predominantly framed in prevention terms and will attract more participants with a prevention-focus. Turn The Tide, on the other hand, opposes “the harsh policies to the right” and “the neoliberal policies of Balkenende II”. Moreover, they are campaigning for “a more progressive policy” (www.keerhettij.nl). Turn the Tide opposes the centre-right government, and its prognosis is a change of government towards a left coalition. We speculate that the collective action frame of Turn the Tide is predominantly framed in promotion terms and therefore will attract more participants with a promotion-focus. To test our speculations, calls for protest participation on the part of the trade union federations and the Turn the Tide alliance were compared on promotion and prevention characteristics. We chose calls for demonstration participation to be compared because these are persuasive communications from the different movements, intended to communicate to the public what the organization stands for. We asked eight employees (one male and seven females) of the social science faculty of the Vrije Universiteit Amsterdam (blind to the hypothesis) to read the following description of prevention and promotion from Shah and Higgins (2001, p. 694): “Regulatory focus theory distinguishes between two distinct systems of selfregulation—promotion and prevention. A promotion-focus relates to nurturance needs, and individuals in such a focus state are more concerned with advancement and accomplishment through the attainment of their hopes and aspirations (i.e., “ideals”). A prevention-focus relates to security needs, and people in such a focus state are more concerned with protection and safety through fulfilling their responsibilities and requirements (i.e., “oughts”)”. We checked whether they understood what a prevention- and promotion-focus implied (all eight indicated they understood it well). Our coders were then requested to read carefully the calls for demonstration participation (length one page) from the trade union federations and Turn the Tide. The coders were asked to indicate whether the respective texts had more prevention than promotion characteristics. Hence, coders could rate both texts as either more promotion or more prevention characteristic. Six out of eight coders rated the Turn the Tide call for the demonstration as possessing more promotion characteristics, whereas six out of eight rated the call for the demonstration from the trade union federations as possessing more prevention characteristics, thus coders agreed 75% of the time. Cohen’s kappa was .60 indicating that the reliability between the coders was substantial (Landis & Koch, 1977). This confirmed our assessment. REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 114 This leads us to assume that the collective action frame of the Turn the Tide alliance will attract promotion- rather than prevention-focused protesters, whereas the collective action frame of the union will attract prevention- rather than promotion-focused protesters. 5.1.5 Hypotheses about regulatory fit and participation motives. In their collective action frames, social movement organizations emphasize different goals, ideologies and solutions. Indeed, our test confirmed that the trade union federations defined their collective action frames in more prevention terms and the Turn the Tide alliance in more promotion terms. We hypothesized that a collective action frame in prevention terms would resonate more for prevention-focused protesters, and one in promotion terms for promotionfocused protesters. Taking this a step further, we suggest that regulatory characteristics of collective action frames will also influence protesters’ participation motives. As the unions’ collective action frame is defined in prevention terms, we assume that it will attract more prevention-focused protesters employing “typical” prevention motives, whereas the collective frame of the alliance, which is defined in promotion terms, will attract promotion-focused protesters employing “typical” promotion motives. As the trade union federations emphasize early retirement rights as one of their goals, we assume that their collective action frame will appeal to prevention-focused protesters who are motivated to participate in collective action by a threat to their security needs. The aim of the protest event staged by the union federations is to remove the collective threat to security needs (i.e., change of the government proposals regarding the budget cuts). This aim may be attainable through collective action participation. Therefore we hypothesize that instrumental motives are more important in the context of prevention-oriented protest rather than promotion-oriented protest. As ideal goals (aspirations, ideals, wishes) trigger a promotion-focus, we expect ideology to be an important motive to participate for promotion-focused participants. Because we expect more promotion-focused people at demonstrations staged by organizations defining their collective action frame in promotion terms, we expect that ideology motives are more important in the context of promotion-oriented protest than in the context of prevention-oriented protest. We hypothesize that both protesters employing instrumental motives and protesters employing ideology motives are angry, although for different reasons. Therefore, we assume that for prevention-focused protesters at the union demonstration and promotion-focused protesters at the Turn the Tide alliance demonstration group-based anger motives are equally important. CHAPTER 5 115 Because oughts (which tend to trigger a prevention-focus) involve duties, obligations and responsibilities that generally are interpersonal, whereas ideals (which tend to trigger a promotion-focus) involve aspirations that are often personal (Pham & Higgins, 2005), we propose that a group norm to participate in collective action (i.e., inner obligation to behave as a “good” group member, cf. Stürmer et al., 2003) will carry more weight for prevention-focused people rather than promotion-focused people. We expect that prevention-oriented protest attracts prevention-focused protesters, therefore we hypothesize that this felt inner obligation, and accordingly identity motives, play a more important role in the context of prevention-oriented protest than in the context of promotion-oriented protest. More specifically, in the context of prevention-oriented protest we hypothesize a mediational role of inner obligation between prevention-focus and identity motives. 5.2 Measures As explained in the introduction to this chapter, we are interested, among other things, in the effect of regulatory fit between regulatory characteristics of collective action frames and the regulatory focus of individual protesters and participation motives. The predictor variables are: promotion- and prevention regulatory focus and instrumental, identity, ideology and group-based anger participation motives. The outcome variable indicates whether someone took part in promotion- or prevention-oriented protest. Regulatory Focus. The regulatory focus was measured with the short version of RFQproverb (Stekelenburg & Klandermans, 2003). Participation motives. To assess instrumental, identity, ideology and group-based anger motives, the same measures as in Chapter 4 were applied. Felt inner obligation. “To what extent did this group influence your decision to participate in this demonstration?” This operationalization of inner obligation served as a proxy for want of something better. We assume that it is a possible substitute for the conceptualization of Stürmer et al., 2003, because with a felt inner obligation the group norm rather than an autonomous individual decision impacts on the motivation to participate. We presume that individuals who are sensitive to this group norm will indicate that their decision to take part in the protest event is influenced by the group, whereas protesters who are not sensitive to this group norm will indicate that their decision is not influenced by the group. REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 116 The outcome variable. This is a categorical variable that indicates that someone took part in promotion-oriented protest (Turn the Tide alliance) = 0, or prevention-oriented protest (union) = 1. 5.3 Results 5.3.1 Preliminary analyses We start with a presentation of the preliminary results differentiated for the two demonstrations (promotion- and prevention-oriented). Table 5.1 shows correlations, means and standard deviations for prevention, promotion, instrumental, identity, ideology, group-based anger and motivation. Higher means indicate more influence of the respective variables. Table 5.1 Correlations, means and standard deviations for prevention, promotion, instrumental, identity, ideology, group-based anger and motivation of two types of protest (promotion-oriented above the diagonal in bold; prevention-oriented under the diagonal). Variables 1 2 3 4 5 6 7 1. promotion -- .07 -.01 .10 .11 -.04 .13 2. prevention .14* -- .05 .06 .07 .08 .10 3. instrumental .06 .14* -- - -.07 .12 .10 .12 4. identity .09 .15* .39** -- .26** .16* 5. ideology .09 -.05 .20** .07 -- 6. group-based anger .23** .02 .28** .08 .34** -- 7. motivation .11 .11 .19** .24** .33** .27** -- 1 2 3 4 .28** .41** .52** 5 .43** 6 7 M promotion-oriented 5.27 4.07a 19.12 a 3.76a 6.43 5.21 a 6.29a M prevention-oriented 5.01 4.33c 23.45c 4.46c 6.40 5.45d 6.51b SD promotion-oriented 0.85 0.94 14.85 2.72 0.77 1.55 1.23 SD prevention-oriented 1.12 15.96 2.15 0.74 1.48 0.83 Note. ab 0.89 * p < .05, ** p < .01 =differ significant at p < .05, ac= differ significant at p < .001, ad =differ at p < .10 CHAPTER 5 117 Table 5.1 reveals that promotion is higher for the promotion-oriented (M = 5.27) than for the prevention-oriented protest (M = 5.01, F = 6.94, p < .001) and prevention is higher for prevention-oriented protest (M = 4.33) than for promotion-oriented protest (M = 4.07, F = 4.39, p = .03). Indeed, the prevention-oriented protest attracted more prevention-focused protesters, whereas the promotion-oriented protest attracted more promotion-focused protesters. Hence, there seems to be a regulatory fit between individual focus and organizational orientation, but did the regulatory fit also influence protesters’ participation motives? We predicted and found that instrumental motives were higher for the unions (M = 23.45) than for the alliance (M = 19.12, F = 8.71, p < .001), indicating that protest events with a collective action frame in prevention terms attract prevention-focused protesters who take the instrumental pathway to protest participation aimed at changing a threat to their (social) security goals. Moreover, instrumental motives are more strongly related to motivation to participate in prevention-oriented protest (r = .19, p < .001) than promotion-oriented protest (r = .12, ns), indicating that instrumental motives influence motivation to participate more strongly in prevention-oriented protest than in promotion-oriented protest. We expected that identity motives would be higher in prevention-oriented (M = 4.46) than promotion-oriented protest events (M = 3.76, F = 24.44, p < .001). This is what we found; it demonstrates that the identity path plays a significantly more important role in prevention- than in promotion-oriented protest. However, identity motives are strongly related to motivation to participate both in promotion-oriented protest (r = .28, p < .001) and in prevention-oriented protest (r = .24, p < .001). Ideology motives are important for protesters at both protests. However, ideology motives are more strongly related to motivation to participate in promotion-oriented protest (r = .52, p < .001) than in prevention-oriented protest (r = .33, p < .001, Fisher r-to-z transformation: z = 2.43, p = .01), indicating that ideology motives influence more strongly motivation to participate in promotion-oriented protest than in prevention-oriented protest, despite being high in both prevention- and promotion-oriented protest. Protesters at the prevention-oriented protest are angrier (M = 5.45) than protesters at the promotion-oriented protest event (M = 5.21, F = 2.78, p = .09). Interestingly, in promotionoriented protest, group-based anger is more strongly correlated to motivation to participate (r = .43, p < .001) than in prevention-oriented protest (r = .27, p < .001, Fisher r-to-z transformation: z = 1.96, p = .05), indicating that, despite being less angry than protesters at the preventionoriented protest, protesters at the promotion-oriented protest are more spurred to participate by their anger than protesters at the prevention-oriented protest—possibly even more so, because a REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 118 comparison of the face-to-face interviews and the surveys (see Chapter 3) revealed that the union protesters were angrier at home whereas the alliance protesters were angrier during the demonstration. Thus, even without the mere presence of other protesters sharing their view, the motivational strength of the alliance protesters is more strongly influenced by anger than that of the unionists. Protesters at the prevention-oriented protest are more strongly motivated to participate in protest (M = 6.51) than protesters at the promotion-oriented protest event (M = 6.29, p = .04). Collective action frames defined in prevention terms seem to attract prevention-oriented protesters, and their motivation is related to “typical” prevention motives, such as instrumental (r = .19, p < .001) and identity motives (r = .24, p < .001). However, contrary to our hypothesis, ideology motives do play an important role in this prevention-oriented protest too (r = .33, p < .001). Collective action frames defined in promotion terms attract promotion-focused protesters and their motivation is related to ideology motives (r = .52, p < .001). Yet, in this promotionoriented protest, protesters are—unexpectedly—spurred by identity motives too (r =.28, p = .01). 5.3.2. Main analyses A social movement organization-participant fit implies that the odds of participating in one demonstration rather than the other can be predicted by regulatory focus. The more preventionfocused individuals are, the more likely they are to take part in the prevention-oriented protest event, and the more promotion-focused they are, the more likely they are to participate in the promotion-oriented protest. Furthermore, we hypothesized that regulatory characteristics of collective action frames would also influence protesters’ participation motives. The more people employ an instrumental and identity motive, the more likely it is that they will take part in prevention-oriented protest; and the more people employ an ideology motive, the more likely it is that they will participate in promotion-oriented protest. In order to test these hypotheses, we regressed the odds that an individual would take part in a promotion-oriented protest event rather than a prevention-oriented protest event on his or her regulatory focus and participation motives in a hierarchical logistic regression analysis. See Table 5.2 for an overview. CHAPTER 5 119 Table 5.2 Hierarchical logistic regression with the odds of participating in promotion- vs prevention-oriented protest. Protest eventa Predictor B SE B eB Model 1 Prevention-focus .34*** .12 1.41 Promotion-focus -.32*** .12 .72 Prevention-focus .29** .12 1.33 Promotion-focus -.32*** .12 .73 Instrumental motive .29** .12 1.34 Identity motive .34*** .12 1.40 -.28* .13 .75 .11 .13 1.12 Model 2 Ideology motive Group-based anger motive Model 1 Model 2 Constant .09 .10 df 2 χ2 15.52*** Nagelkerke R2 .06 Hosmer and Lemeshow χ2 18.82* % correctly classified 63% Note. 4 35.95*** .13 14.23 63% B = unstandardized beta coefficient; SE = standard error; eB = exponentiated B= odds ratio11 a = Scored: promotion-oriented protest = 0, prevention-oriented protest = 1 *p < .05. **p < .01. ***p < .001. In the first step we entered prevention- and promotion-focus, and in the second step the instrumental, identity, ideology and group-based anger participation motives were entered. The analysis revealed that, overall, the differences between the two demonstrations were significant 11 The exponentiated B value indicates the extent to which the probability that someone will take part in preventionoriented protest increases (if the value is above 1) or decreases (if the value is below 1) by the amount of the value times the original probability. REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 2 ( 120 (6; 449) = 35.95, p < .001). The odds of participating in one demonstration rather than the other can be predicted by regulatory focus: the more prevention-focused people are, the more likely it is that they will participate in a prevention-oriented protest event (prevention-focus Wald’s 2 (1; 449) = 8.77, p < .001) and the more promotion-focused people are, the more likely it is that they will take part in the promotion-oriented protest event (promotion-focus Wald’s 2 (1; 449) = 7.71, p < .001). The odds of participating in the prevention-oriented protest increase by 41% if the strength of prevention-focus increases, while the odds of participating in preventionoriented protest with a stronger promotion-focus decreases by 28%. Do the regulatory characteristics of collective action frames also influence the protesters’ participation motives? To test this assumption, the motives were entered in the second step. By entering the motives into the equation, the explained variance increases from .06 to .13, while the contribution of prevention- and promotion-focus scores to the prediction as to which demonstration protesters will take part in does not substantially change (prevention-focus Wald’s 2 (1; 449) = 5.82, p = .02 and promotion-focus Wald’s 2 (1; 449) = 7.00, p < .001). This suggests that the regulatory fit effect remains significant, including when the participation motives are entered into the equation. The contribution of instrumental, identity and ideology participation motives to the prediction that protesters would participate in one protest event rather than the other was significant and in the expected direction. A stronger instrumental motive increases the odds that someone will take part in prevention-oriented protest (Wald’s 2( 1; 449 = 5.52, p = .02). Likewise, stronger identity motives increase the odds that someone will take part in prevention-oriented protest (Wald’s 2 (1; 449) = 7.99, p < .001). Ideology motives seem to have the opposite effect, indicating that a stronger ideology motive decreases the odds that people will take part in the prevention-oriented protest event (Wald’s 2 (1, 449) = 4.42, p = .04). As hypothesized, group-based anger motives do not differentiate between participants in protest events. This suggests that indeed the promotion-oriented protest attracted promotion-focused protesters pushed by ideology and group-based anger motives, whereas the prevention-oriented protest attracted prevention-focused protesters pushed by instrumental, identity and group-based anger motives. 5.3.3 Does inner obligation mediate between prevention-focus and identity motives? We assumed, and our findings suggest, that prevention-oriented protest attracted preventionfocused protesters pushed by (among other things) identity motives. This hypothesis is based on the assumption that prevention-focused protesters are more sensitive to the normative pressures CHAPTER 5 121 to participate, in other words, a felt inner obligation to participate. Moreover, we expect that a felt inner obligation strengthens the correlation between prevention-focus and identity motives. To test this role of felt inner obligation on prevention-focus and identity, we conducted mediational analyses for the prevention- and the promotion-oriented protest following Baron and Kenny’s (1986) three-step procedure. Table 5.3 gives an overview of the three regressions in the context of prevention-oriented protest. First, we regressed the mediator (inner obligation) on the predictor (prevention-focus). The result indicates that the first prerequisite for mediation is met ( = .21, p < .001). Next, we regressed the outcome variable (identity motives) on the predictor variable (prevention-focus). The result provides support for the second prerequisite ( = .17, p < .02). Finally, we regressed the outcome variable (identity motives) on the mediating variable (inner obligation), with the predictor variable included in the equation. The beta weight for inner obligation was significant, indicating support for the third prerequisite. Most important, when prevention-focus and inner obligation were included as predictors of identity motives, inner obligation predicted identity motives significantly, from = .17 to = .65, t(215) = 12.77, p < .001, whereas prevention-focus was reduced = .03, t(215) = .60, p = .55; this indicates evidence for full mediation (Baron & Kenny, 1986). Table 5.3 Three regressions to test mediation of inner obligation on prevention-focus and identity motives in prevention-oriented protest: (1) inner obligation on prevention; (2) prevention on identity motives; and, (3) identity on inner obligation and prevention. Prevention Equation 1 Inner obligation Equation 2 Prevention Identity .21*** .17* Equation 3 Prevention Inner obligation 10.13*** 6.63* (1; 228) (1; 227) .04 .03 *p < .05. **p < .01. ***p < .001. Model F df R2 Note. Identity .03 .65*** 87.24*** (2; 225) .44 Inner obligation reduced the effect of prevention-focus on identity motives significantly (Sobel zvalue = 3.12, p < .001, direct: -.03, indirect: .14). Thus, the fact that stronger prevention-focus REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 122 enhances identity motives in prevention-oriented protest can be completely explained by an inner obligation to behave as a “good” group member. Why do prevention-focused people want to behave as “good” group members? We hypothesized and found that prevention-focused people, due to their sensitivity to duties, obligations and responsibilities (Higgins, 1997; 1998) and endorsement of conformity values (Van-Dijk and Kluger, 2004; see also Chapter 2) would be sensitive to the norm of participation in social movement organizations. Figure 5.1 visualizes this mediational analysis. .17** Prevention (.03) .21** Inner obligation Identity motives .66*** (.65***) Figure 5.1. Mediational model for inner obligation on prevention-focus and identity motives in prevention-oriented protest. Does this medational role of inner obligation also hold for promotion-focus and identity motives? Again, we followed Baron and Kenny’s (1986) three-step procedure to test this mediational role. This analysis revealed no mediational effect of inner obligation on the relation between promotion-focus and identity in the context of prevention-oriented protest (Sobel z-value = 1.16, p = .25). Thus, in contrast to a prevention-focus, an inner obligation in prevention-oriented protest does not mediate between promotion-focus and identity motives. We assumed a mediational role of inner obligation between prevention-focus and identity motives in prevention- rather than promotion-oriented protest; but what about promotion-oriented protest, would a similar mediation process be present in that mobilizing context? The same threestep procedure revealed no mediational effect (Sobel z-value = .30, p = .76), indicating that it is only in the context of prevention-oriented protest that inner obligation mediates between prevention-focus and identity motives. The mediation of inner obligation between prevention-focus and identity motives might be due by a fit between the individual self-regulatory focus and the regulatory orientation of the collective action frame of the organization. In other words, would a stronger promotion-focus similarly enhance an inner obligation to participate in the context of promotion-oriented protest? To answer this question we tested a mediation model with inner obligation as mediator on CHAPTER 5 123 promotion-focus and identity motives in the context of promotion-oriented protest. This mediational analysis revealed no mediational effect (Sobel z-value = 1.13, p = .26), suggesting that inner obligation does not mediate between promotion-focus and identity motives in the context of promotion-oriented protest. 5.4 Discussion In Chapter 4 we operationalized social movement context by employing Turner and Killian’s (1987) concept of action orientation. In this chapter, others, who were unaware of the research questions, determined whether a call for demonstration was more prevention-oriented or more promotion-oriented. The collective action frame of the union was assessed as preventionoriented, whereas the collective action frame of the Turn the Tide alliance was assessed as promotion-oriented. Our study suggests that frame alignment may generate a regulatory organization-participant fit; that is, the more an organization emphasizes prevention aspects in its campaign, the more it attracts prevention-focused protesters; and the more it emphasizes promotion aspects the more it tends to attract promotion-focused protesters. Moreover, regulatory fit seems to affect participation motives as well. Prevention-oriented protest tends to attract prevention-focused protesters displaying “typical” prevention motives, such as identity and instrumental motives aimed at social security, whereas promotion-oriented protest tends to attract promotion-focused protesters employing “typical” promotion motives, such as ideology motives. By identifying the regulatory characteristics of social movement contexts and selfregulatory foci we were able to show that prevention-oriented protesters are sensitive to the group norm of participation in prevention-oriented collective action. Organizations staging prevention-oriented protest tend to be more attractive to prevention-focused protesters who display strong identity motives due to a felt inner obligation to behave as a “good” member of the social movement organization staging the prevention-oriented protest. Such a norm of participation is, however, only influential for prevention-focused people in the context of prevention-oriented protest. Some caution is advised, because our measure of inner obligation is not ideal. All this does not yet demonstrate that regulatory focus has a steering function in the individual motivational mechanisms that lead to protest participation. Indeed, we still do not know whether self-regulatory processes at the individual level shape participation motives. Three routes to protest participation have been proposed so far, but there is no elaboration on which route will prevail for whom, and why. The answer to these kinds of questions presupposes a REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 124 steering mechanism. In addition to our three possible paths to collective action, we conceive of regulatory focus as a moderating variable that can explain why one path to protest participation will prevail over another. In the remainder of this chapter we will elaborate on this matter. 5.5 Does regulatory focus moderate the paths taken to protest participation? In Chapter 4 we showed that participants in the context of power-oriented protest events are more inclined to take the instrumental/identity route with an ideological component to protest participation, whereas participants in value-oriented protest events are inclined to take the ideology route. In the previous section of this chapter, we showed that prevention-oriented protest tends to attract prevention-focused protesters applying typical prevention motives, whereas promotion-oriented protest tends to attract promotion-focused protesters employing typical promotion motives. The question remains as to what extent this is a fully mechanical, involuntary process, that is to say, a matter of a match of characteristics of the environment and characteristics of an individual, or rather a matter of choice, that is to say, do individuals have a specific psychological make-up that steers them onto one of the possible routes to collective action rather than another? If it is the latter, what mechanism could possibly steer an individual onto one path rather than another? We hold that regulatory focus could be such a steering mechanism. In Chapter 1 we elaborated on the role of regulatory focus. In this chapter we test some hypotheses derived from the reasoning developed. Structural equation modelling will be used to test a general model of protest participation developed on the basis of these hypotheses. Subsequently, this general model will be adapted to different social movement contexts. We will employ the distinction in regulatory characteristics following from the study described in the preceding part of this chapter to differentiate between movement contexts. Finally, applying multiple-group analysis, we will test whether the models of the prevention- and promotion-oriented protest differ from each other. In other words, does the motivational pattern in a prevention-oriented protest event differ substantially from the motivational pattern in a promotion-oriented protest event? 5.5.1 Regulatory focus and participation motives. Self-regulation can be seen as an adjustment strategy of the self to a changing social and political environment. Higgins (1997; 1998) argues that self-regulation can be pursued either with means CHAPTER 5 125 that are approach-oriented or with means that are avoidance-oriented. People employ approach-oriented promotion-focused self-regulation strategies when they approach pleasure, and avoidance-oriented prevention-focused strategies when avoiding pain. These two modes of selfregulation are the source of fundamentally different goals, needs and values. Given the prioritization of fundamentally different goals, needs and values, we assumed that regulatory focus also influenced the reason(s) why people were motivated to participate in political protest. In statistical terms, we hypothesized that regulatory focus moderated the paths taken to protest participation. In Chapter 1 we formulated hypotheses regarding self-regulation and participation motives. In this section, we will test some of these hypotheses. To do so, we will first extend the general model of protest participation tested in Chapter 4 by adding regulatory focus, followed by an adaptation of this extended model to the different mobilizing contexts. The baseline model in Chapter 4 revealed three entrances: instrumental, identity and ideology motives, but how does regulatory focus control which path is taken? We hypothesized that promotion- and prevention-focused protesters with an instrumental participation motive wanted to solve diverging problems by protest participation. We hypothesized that prevention-focused people would be motivated to participate in political protest for instrumental reasons when security needs were threatened and protection and responsibility goals were obstructed, whereas promotion-focused people would be motivated when nurturance needs were threatened and growth and advancement goals were obstructed. Because both demonstrations were so clearly framed around social security issues, we hypothesized that prevention- rather than promotion-focused protesters would take an instrumental pathway to this protest. In the previous section of this chapter we showed that regulatory focus controls group identification processes. Prevention- rather than promotion-focused people are inclined to employ identity motives due to a felt inner obligation to behave as a “good” member in preventionoriented protest. Therefore, we propose that a group norm to participate in collective action (i.e., employ an identification participation motive) will carry more weight for prevention-focused than promotion-focused people. The ideology path to protest participation refers to people’s ideology or values and the assessment that these ideologies or values have been violated. Prevention- and promotionfocused people prioritize values differently however, and therefore we assume that motivation to take the ideology path to political protest stems from violation of divergent values: preventionfocused people are motivated to take the ideology pathway to collective action when tradition, REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 126 conformity or security values have been violated, and promotion-focused people when selfdirected and stimulation values have been violated. Given the different value prioritization by promotion- and prevention-focused people, we assume that progressive political ideology is appealing to promotion-focused people and conservative political ideology to prevention-focused people. Turn the Tide emphasized the progressive political ideology behind their aims. Therefore we hypothesize that promotion- rather than prevention-focused protesters take an ideology pathway. 5.6 Measures We employ the same measures as in the previous section of this chapter. We refer therefore to section 5.2 for further information on measures. 5.7 Results To test the models that reflect our arguments, we employed the Analysis of Moment Structures (AMOS) software (Arbuckle, 1997). The models integrate our main ideas about how motivational patterns differ as a function of the self-regulatory focus of the protester and social movement context. We will build the model starting with the baseline model proposed in Chapter 4. First, we will extend this baseline model by adding regulatory focus. Thereafter, a multiplegroup analysis will be conducted to test the differences in regulatory focus of protesters and their participation motives in the context of promotion- and prevention-oriented protest. 5.7.1 SEM: Baseline model of protest participation steered by regulatory focus. We examined a model that represents our argument. The predictive model integrates our main hypotheses, predicting that a prevention- rather than a promotion-focus steers an individual onto the instrumental and identity paths, whereas a promotion- rather than a prevention-focus steers an individual onto the ideology path. Figure 5.2 displays the hypothesized pathways and the related betas. .12** identity .20*** .14*** prevention .10* promotion anger .15*** .13*** .08º .14*** instrumental .35*** .31*** .22*** motivational strength CHAPTER 5 127 Note. º = p < .10, * = p < .05, * * = p < .01, *** = p < .001 Figure 5.2. SEM: baseline model accounting for motivational strength steered by regulatory focus. This hypothesized model fits the data very well ( 2 (9; 449) = 8.5, p = .48, CFI = 1.00, NFI = .97, RMSEA = < .001, and the Squared Multiple Correlation of motivation = .26). Thus, by adding regulatory focus to the baseline model that tested the relative weight of the four participation motives (see Chapter 4), the explained variance remains the same. This is actually what we would expect. After all, regulatory focus is included to explain which motivational pattern will prevail rather than to increase the variance explained. Prevention-focus does indeed steer an individual onto an instrumental and identity path, whereas promotion-focus steers an individual onto an ideology path. Admittedly, the betas are rather weak, especially for promotion-focus, though (marginally) significant. Furthermore, the Squared Multiple Correlation of instrumental, identity and ideology motives is .10, indicating that 10% of the variance in the participation motives is explained by regulatory focus. However, the observed pattern is in the expected direction. Prevention-focused protesters are inclined to take the instrumental ( = .10, p = .04) rather than the ideology route ( = .02, p = .60) to protest participation, whereas promotion-focused people are inclined to take the ideology ( = .08, p = .07) rather than the instrumental ( = .02, p = .66) route. Moreover, the identity path is more frequently taken by people with a prevention-focus ( = .12, p = .01) than with a promotion-focus (( = .07, p = .28). Finally, prevention- and promotion-focus are not directly related to groupbased anger and do not directly influence the strength of the motivation to participate. Regulatory focus seems indeed to moderate the paths taken to protest participation: the stronger the prevention-focus, the more protesters are steered onto the instrumental or identity path rather than an ideology path; and the stronger the promotion-focus, the more protesters are steered onto an ideology rather than an instrumental or identity path. In the previous part of this chapter we showed that inner obligation mediates on prevention- rather than promotion-focus and identity. The structural equation model shows the same results in a single analysis: a direct effect from prevention-focus to identity and no such effect for promotion-focus. This suggests that REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 128 prevention- rather than promotion-focused protesters are sensitive to the norm of participation in social movement organizations. Plausible alternative models. Although the reported path analyses provided strong support for the hypothesized routes to protest participation, it is possible that other models may fit the data equally well, or better. Therefore, we tested two possibly competing models. First, we reasoned that promotion-focus might explain the instrumental path, and prevention-focus the ideology path. Thus, we reversed the instrumental and ideology routes steered by the two regulatory foci. This model fitted poorly to the data ( 2 (9; 449) = 20.40, p = .02, CFI = .95, NFI = .92 and RMSEA = .06). The significant chi-square indicates that this hypothesized alternative model does not adequately explain the observed pattern of data. Moreover, the pathways from promotion-focus to instrumental ( = .02, p = .66) and from prevention to ideology ( = .02, p = .60) proved to be unrelated. This suggests that promotion-focused people are much less inclined to take an instrumental path to protest participation, whereas prevention-focused people are much less inclined to take an ideology path to protest participation. Second, we reasoned that prevention- and promotion-focused people were equally likely to take both routes; put differently, regulatory focus does not distinguish between pathways to protest events. Although this model showed a good fit ( 2 (7; 449) = 7.81, p = .35, CFI = 1.00, NFI = .97 and RMSEA = .02), the pathways from promotion to instrumentality and from prevention to ideology remained insignificant ( = .01, p = .87 and = .01, p = .76, respectively). Thus, promotion-focused people are indeed more inclined to take the ideology rather than the instrumental route, whereas prevention-focused people are, in fact, more inclined to take the instrumental rather than the ideology route to protest participation. This suggests that regulatory focus indeed controls which path is taken to protest participation. Adding regulatory focus as the steering mechanism to our theory of protest participation provides an answer to the question: “Who participates for what reason?” The question remains as to whether the reasons why people participate are contingent upon the social movement context. We propose that a prevention-focus steers people onto an instrumental and identity path in the context of prevention-oriented protest, whereas a promotion-focus steers people onto an ideology path in the context of promotion-oriented protest. A multiple-group analysis will be conducted to test this idea. 5.7.2 Multiple-group analysis: regulatory focus and context. In an attempt to compare the motivational patterns for prevention- and promotion-oriented protest, we will apply the same multiple-group analyses as in Chapter 4 proposed by Byrne CHAPTER 5 129 (2004). The presented and tested baseline model will accordingly be tested for the prevention- and promotion-oriented protest separately. Following an inductive approach, the insignificant paths will be fixed to zero. As indicated, we hypothesize that an ideology path is irrelevant in the explanation of the variance in motivational strength in the context of preventionoriented protest, whereas an instrumental and identity path are of minor importance in the context of promotion-oriented protest. Hence, we hypothesize that different paths will be insignificant, contingent on social movement context. Finally, the unconstrained and the constrained models will be compared. If the constrained model fits the data as well as the unconstrained model, the constrained model is to be preferred for parsimony reasons. Baseline model submitted to prevention- and promotion-oriented protest events. We hypothesized that the motivational patterns for the prevention- and promotion-oriented protest events would differ from each other. To test this hypothesis, the baseline model (see Figure 5.2) will be submitted to a test using the two distinct samples. This offers the opportunity to test whether a prevention-focus steers people onto an instrumental and identity path in the context of prevention-oriented protest, whereas a promotion-focus steers people onto an ideology path in the context of promotion-oriented protest. In Chapter 4 we showed also that group identification shapes what people are angry about. Therefore, we hold that the mediational role of instrumental and ideology motives between identity and group-based anger motives will be different for promotion- and prevention-oriented protest. Put differently, we hypothesize that, in the context of prevention-oriented protest, instrumental rather than ideology motives mediate between identity and group-based anger motives, whereas, in the context of promotion-oriented protest events, ideology rather than instrumental motives mediate between identity and group-based anger. Thus, in preventionoriented protest events, prevention-focused protesters will employ identity motives, which will strengthen their instrumental motives, which in turn will make them angrier, whereas, in promotion-oriented protest events, promotion-focused protesters will employ identity motives, which will strengthen their ideology motives, which in turn will make them angrier. Figure 5.3 provides an overview of the baseline model submitted to the prevention-oriented protest (above) and the promotion-oriented protest (below). identity .12* .41*** prevention .09º .06 promotion instrumental .18*** .18*** .20*** .30*** anger .13* .20*** motivational strength REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 130 Note. = p < .10, * = p < .05, *** = p < .001. Dashed lines imply insignificant or negative cross-loadings. Figure 5.3. Baseline model submitted to prevention (above) - and promotion-oriented protest (below) separately. The baseline model adequately fits the data in both samples (prevention-oriented protest: 2 (9; 240) = 15.49, p = .09, CFI = .95, NFI = .90 and RMSEA = .05, and Squared Multiple Correlation [R2] of motivation = .18; promotion-oriented protest: 2 (9; 209) = 12.67, p = .18, CFI = .97, NFI = .92 and RMSEA = .04, and Squared Multiple Correlation [R2] of motivation = .33). The baseline model tested for the separate samples shows different insignificant paths: in the prevention-oriented protest, the path from identity to ideology failed to reach significance. In the promotion-oriented protest sample, on the other hand, paths from prevention to instrumental motives and identity motives, ideology and identity to instrumental motives, and instrumental to group-based anger motives failed to reach significance. Note that, as hypothesized, motivational strength to participate in promotion-oriented protest is influenced by the ideology path controlled by promotion-focus, whereas in the context of prevention-oriented protest an instrumental path controlled by prevention-focus influences the motivational strength. However, contrary to our CHAPTER 5 131 expectations, promotion-focused protesters are inclined to take the ideology route in prevention-oriented protest. Submitting the baseline model to a test for the two contexts reveals diverging motivational patterns for the two settings. However, of most interest concerning our question about regulatory focus as moderator contingent on social movement context is the finding that the motivational patterns of the prevention- and promotion-oriented protest events differ: an ideology path controlled by promotion prevails in the context of promotion-oriented protest and an instrumental and identity path controlled by prevention-focus and an ideology controlled by promotion-focus prevails in the context of promotion-oriented protest. In Chapter 4 the models of hierarchical regressions and structural equation modelling revealed that, in the context of power-oriented protest, the ideology route played a major role in the motivational pattern. This finding was contrary to our expectations, because we expected that the instrumental route would prevail in the context of prevention-oriented protest. Interestingly, instrumental motives are part of this ideology path steered by promotion-focus, suggesting that promotion-focused protesters take the ideology route in prevention-oriented protest not only because they want to express their view and vent their anger, but also because they want to solve the problem. This is an intriguing finding to which we will return in the discussion. Hence, implementing a steering mechanism in our model offers the opportunity to explain which protesters took the ideology route to prevention-oriented protest. The data suggest that promotion- rather than prevention-focused people are inclined to take the ideology route to prevention-oriented protest (promotion-focus to ideology is ideology is = .08, p = .10, prevention-focus to = -.06, p = .31). According to the regulatory fit hypothesis, prevention- rather than promotion-focused people are attracted to prevention-oriented protest. Why, then, were promotion-focused people taking part in a prevention-oriented protest? We will reflect on this question in the discussion. These structural equation models replicate the findings from the previous part of this chapter where we hypothesized, and demonstrated by means of mediational analyses, that the psychological process underlying identification processes depends on social movement context and individual regulatory processes. In the previous part of this chapter, mediational analyses revealed that inner obligation mediates between prevention-focus and identity motives, but only in the context of prevention-oriented protest. The structural equation models for prevention- and promotion-oriented protest suggest the same. In the context of prevention-oriented protest, prevention-focus and identity motives are significantly related ( = .12, p = .02), whereas the two are not significantly related in the context of promotion-oriented protest ( = .07, p = .44). REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 132 Promotion-focus, on the other hand, is not significantly related to identity either in the context of promotion-oriented protest ( = .09, p = .24) or in the context of prevention-oriented protest ( = .05, p = .32). This implies that the norm to behave as a “good” group member carries more weight for prevention- than promotion-focused protesters, but only in prevention-oriented protest. Thus, both mediational analyses and structural equation modelling reveal that group identification processes are controlled by regulatory processes at the individual level and at the level of the social movement context. It is a paradoxical finding that identity motives have a direct effect on motivational strength in the context of promotion-oriented protest. We hypothesized and found that promotion-focused people were inclined to take the ideology route in promotion-oriented protest. Moreover, we hypothesized and found that promotion-focused people were not inclined to take the identity path to protest. This suggests that identity motives would not influence motivational strength in the context of promotion-oriented protest. But they do, although neither promotionnor prevention-focus steers an individual onto the identity path in the context of promotionoriented protest. It seems that promotion-focused people identify with social movement organizations staging promotion-oriented protest, but that this does not influence their decision to participate or not. 5.7.3 SEM: modified model of participation in prevention-oriented protest events. Following an inductive approach, the next step in the multiple-group analysis will be to force the non-significant paths to zero. The following paths will be forced to zero: in the context of prevention-oriented protest, the path from identity to ideology motives; and, in the context of the promotion-oriented protest, the paths from prevention-focus to instrumental motives, preventionfocus to identity motives, identity to instrumental motives, ideology to instrumental motives, and instrumental to group-based anger motives. The modified model of the prevention-oriented protest event showed adequate fit ( 2 (10; 238) = 16.35, p = .09, CFI = .95, NFI = .90 and RMSEA =. 05, and the Squared Multiple Correlation [R2] of motivation in this model is .17). Figure 5.4 displays the pathways and their betas. identity .12* .18*** .41*** prevention .09º 0 instrumental .20*** anger .13** promotion .18*** .08º ideology .30*** .20*** motivational strength CHAPTER 5 133 Note: º = p < .10, * = p < .05, * * = p < .01, *** = p < .001 Figure 5.4. Modified model submitted to prevention-oriented protest steered by regulatory focus. Then, what model fits the data of prevention-oriented protest better, the baseline model or the modified model whereby the identity/ideology path is fixed to zero? Following Byrne (2004), we will apply the same method as in Chapter 4 to test the chi-square difference by comparing the chi-square of the baseline model submitted to prevention-oriented protest with the modified model. This chi-square difference is 2 ( df. = 1; 240) = .86, p > .50. This chi-square difference is not significant, indicating that the baseline model and the modified model have comparable fit to the data. However, for parsimony reasons the modified model is to be preferred. Thus, a model with an instrumental/identity path steered by prevention-focus and an ideology path steered by promotion-focus, whereby identity motives are not related to ideology, fits the data of a prevention-oriented protest better than the baseline model. Plausible alternative models. We tested two alternative models, namely the “pure” ideology and the “pure” instrumental pathways to protest participation. For the “pure” ideology pathway, the paths representing the instrumental pathway were assigned a “0” to the crossloadings, whereas for the “pure” instrumental pathways the paths representing the ideology pathway were assigned a “0” to the cross-loadings. Both models showed a poor fit to the data (the “pure” ideology model: 2 (16; 240) = 88.14, p < .001, CFI = .44, NFI = .44 and RMSEA = .14, and the “pure” instrumental model: 2 (14, 238) = 65.48, p < .001, CFI = .60, NFI = .58 and RMSEA =. 13). The significant chi-square and the other fit indices indicate that these alternative models do not adequately account for the observed pattern of the data. This suggests that protesters participating in prevention-oriented protest events are not inclined to take a “pure” ideology or a “pure” instrumental pathway. 5.7.4 SEM: modified model of participation in promotion-oriented protest events. In order to test whether indeed the ideology pathway predominates over the instrumental pathway in the context of promotion-oriented protest, we assigned the cross-loadings representing the REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 134 instrumental pathway to “0”. The modified model for promotion-oriented protest fits the data 2 ( (14; 209) = 18.92, p = .18, CFI = .97, NFI = .90 and RMSEA =. 04, Squared Multiple Correlation [R2] of motivation = .34). Moreover, the data reveal that promotion-focused people are inclined to take the ideology rather than the instrumental path to promotion-oriented protest events. Figure 5.5 depicts this model for promotion-oriented protest. identity .0 .0 prevention .14* .0 instrumental .0 anger .30*** motivational strength promotion .22*** .0 .42*** .44*** .09º ideology Note. = p < .10, * = p < .05, * * = p < .01, *** = p < .001. Figure 5.5. Modified model submitted to promotion-oriented protest steered by regulatory focus. Does the modified model with an ideology route fit the promotion-oriented protest data better than the baseline model submitted to the promotion-oriented protest? The change in chi-square of the baseline model submitted to promotion-oriented protest compared with the modified model is 2 ( df. =5; 209) = 6.25, p > .30. This chi-square difference is not significant; this indicates that the baseline model tested in the context of promotion-oriented protest and the modified model have comparable fit to the data, and for parsimony reasons the modified model is to be preferred. Thus, a model accounting for an ideology path controlled by promotion-focus fits the data of a promotion-oriented protest better than a model accounting for both an ideology route controlled by promotion-focus and identity and instrumental paths controlled by preventionfocus. Plausible alternative model. Although the path analysis provided support for the reasoning that promotion-focused protesters take the ideology route to promotion-oriented protest, it is possible that other models fit the data too, or better. One alternative model we tested was the idea that prevention-focused protesters would take an instrumental route to protest 2 participation. This model showed a poor fit to the data ( (14; 207) = 99.66, p < .001, CFI = .33, CHAPTER 5 135 NFI = .36 and RMSEA = .17). The significant chi-square and the other fit measures indicate that this alternative model does not adequately explain the observed pattern of data. This suggests that promotion-focused protesters are not inclined to take the instrumental pathway to promotionoriented protest. Thus, in line with our hypothesis, promotion-focused people are inclined to take the ideology rather than the instrumental route to promotion-oriented protest events. In summary, the modified models for the prevention- and promotion-oriented protest events are substantially different from each other. In the context of promotion-oriented protest, only the ideology path is taken by promotion-focused people, contrary to prevention-oriented protest where prevention-focused people take the instrumental/identity route and promotionfocused people the ideology route. Adapting the baseline model of protest participation to prevention- and promotion-oriented protest yields improvement of the models. More important for our steering function hypothesis, however, is the finding that a prevention-focus steers an individual onto the instrumental/identity path, whereas a promotion-focus steers an individual onto the ideology path. This suggests that regulatory focus configures a frame to interpret the social and political world, and therefore moderates the paths protesters take to protest participation. 5.8 Discussion To our knowledge, this is the first social psychological study of protest participation attempting to explain which pathway to protest participation will prevail for whom, and why. We proposed three possible routes to collective action, and regulatory focus as the mechanism that moderates which route will be taken. Regulatory focus, in our view, helps to explain why one route to protest participation is more likely than others. Regulatory focus as a steering mechanism offers the opportunity to answer the question of why one protester is more inclined to take the instrumental or identity pathway to participation whereas another takes the ideology pathway. Prevention-focused people appear to be inclined to take the instrumental route when their ought goals are obstructed, or the identity route when they feel an inner obligation to behave as a “good” group member. Promotion-focused people, on the other hand, appear to be inclined to take the ideology route because their values have been violated. Our findings indicate that motivational patterns differ as a function of social movement context and suggest explanations why. The collective action frame of Turn the Tide emphasized the erosion of the social system as diagnosis, and comprised a plea for progressive policies as prognosis. The demonstration organized by Turn the Tide attracted promotion-focused protesters REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 136 who then took the ideology route to protest. This suggests that promotion-oriented protest attracts promotion-focused protesters aiming at promoting social change. The collective action frame of the union was also defined around the erosion of the social system, but emphasized erosion of early retirement rights as its focal point. The union demonstration attracted both prevention-focused protesters who took the instrumental and the identity route, and promotionfocused protesters who took the ideology route. This seems to imply that the union attracted prevention-focused protesters who perceived the proposals around early-retirement rights as an obstruction of their security goals and promotion-focused protesters who perceived the social erosion as a violation of their values. The findings that a prevention-focus steers an individual onto the instrumental path seems to correspond to the conclusions of van der Veen (1992) who examined the relation between action goal and action preparedness for unionists. She showed that the action preparedness of unionists of the Christian Trade Union Federation was stronger when goals touched their subsistence security (i.e., security goals and needs), and that action preparedness was stronger when the goal related to already existing rights, in other words, to preventing social change. Adding regulatory focus as a psychological mechanism helps to understand the unexpected finding of an ideology route in prevention-oriented protest. The data reveal that promotion-focused protesters took part in the unions’ prevention-oriented protest through the ideology route. Why does the unions’ prevention-oriented protest attract promotion-focused protesters? The study of De Witte (1995) helps us to understand why. In his study on the entrance motives of unionists, he shows that instrumental motives (expected financial support in the event of strikes and legal aid in the event of conflicts with the employer) prevail, followed by ideology (the social meaning and ideology of the unions). Social motives (the influence of the social surroundings, e.g., becoming a member because of social pressure or expectations of colleagues, family, etc.) are less important and seem to have declined over the last decennia (van Rij, 1995). It may be that people with ideology entrance motives (i.e., recruitment and group attractivity with promotion-focused characteristics) are promotion-focused protesters taking the ideology route. Of course these are speculations; however, it is important for our finding of the ideology path steered by promotion-focus that studies on motives for becoming a member of a trade union also reveal instrumental and ideology routes to trade union membership. Why does the ideology motivational pattern in prevention-oriented protest encompass an instrumental motive, whereas it does not in promotion-oriented protest? A tentative answer may be that the collective action frame of the union emphasized, besides instrumental goals (i.e., early retirement rights), their political efficacy as an organization. It may be that the instrumental CHAPTER 5 137 motive for promotion-focused protest is influenced by the expectancy rather than the value of the early retirement rights; this, we assume, is a security goal related to a prevention-focus, especially as the union stressed its political efficacy much more than Turn the Tide12. Another explanation, related to political efficacy, is that the demonstration was more successful for the trade union federations than Turn the Tide, for three reasons. First, the turnout for the trade unions was enormous, 250,000 as against 50,000 for Turn the Tide; next, soon after the demonstration the government decided to withdraw the most “painful” cutbacks such as early retirement which had been emphasized by the trade union; and, third, the media coverage seemed to focus on the trade union federations, while neglecting Turn the Tide. All these issues together might empower the union protesters more (Drury & Reicher, 1999; 2005) than the Turn the Tide protesters. Empowerment might accordingly have influenced expectations. The more protesters are empowered before, during and after the protest event, the more they expect that the goal of the action may be reached and the higher their expectancies will be. 12 Indeed, promotion and expectancy are more strongly (although not significantly) related to each other in the prevention-oriented protest (r = .10, p < .15) than in the promotion-oriented protest (r = -.02, ns, Fisher r-to-z transformation: z = 1.23, p = .21), while value of the security goal and promotion are not related to each other in either protest event (both r = .00, ns). REGULATORY FOCUS AND POLITICAL PROTEST PARTICIPATION 138 As indicated, these are speculations; future research could disentangle the effects of expectancy and empowerment. Group identification processes are known to influence protest participation but our study refines and augments the understanding of identification processes by introducing regulatory processes. More specifically, by identifying the regulatory characteristics of the campaign and self-regulatory foci we were able to show that only prevention-oriented protesters are sensitive to the group norm of participation in prevention-oriented collective action. Indeed, organizations staging prevention-oriented protest are more attractive to prevention-focused protesters who employ strong identity motives due to an inner obligation to behave as a good member. Apparently, this norm to behave as a good group member is for promotion-focused people less stringent. If it is not an inner obligation to behave as a “good” group member, then, what is it that moves promotion-focused people to take to the streets? We have argued that maintaining moral integrity and emotional coping (i.e., ideology and group-based anger motives) incite an inner obligation as well. Maintaining moral integrity and emotional coping may incite an inner moral obligation to oneself, versus the inner social obligation to other group members incited by group identification. We will return to the distinction between moral and social obligation in the general discussion. Then, why would this felt inner obligation only function in prevention-oriented protest? We suggest that this may be related to the kind of organizations staging the different protest events, the trade union federations and Turn the Tide. The missions, goals and ideology of the two organizations are different and might accordingly attract people with different regulatory characteristics. We assumed that prevention-focused people would identify with “safe” groups that provide security, rather than “upwardly mobile” groups promising growth. The union may be a good example of a safe group in its function of pressure group and may accordingly attract more prevention- than promotion-focused people. Because prevention-focused people are more sensitive to the norm of participation, we assume that this functions more strongly in the union protest event than in the Turn the Tide alliance protest event. The beta weights explaining the relation between the respective regulatory foci and the proposed routes to protest participation vary between .09 and .12, suggesting that regulatory focus explains to a certain extent why people are inclined to take one route rather than another, but also suggesting that that there is room for improvement. Despite the fact that the pattern was in the expected directions in all the studies, regulatory focus seems to be only one of the mechanisms that explains, at least in our studies, a little variance in the employed motives (10%), CHAPTER 5 139 but definitely not all. In particular, the relation between promotion-focus and ideology motives is rather weak (between .08 and .09). We hope to have shown in this chapter that the integration of regulatory focus theory and social psychological approaches to protest participation can be fruitful. We demonstrated in two different ways how regulatory focus theory may improve models that explain protest participation. First, we showed a regulatory fit between social movement context and individual protesters. In other words, collective action frames defined in prevention terms appear to attract prevention- rather than promotion-focused people employing “typical” prevention motives, whereas collective action frames defined in promotion terms attract promotion- rather than prevention-focused protesters employing “typical” promotion motives. Next, we showed that regulatory focus appears to function as a steering mechanism: prevention-focused people are inclined to take the instrumental/identity rather than the ideology route, whereas promotion-focused people are more inclined to take the ideology route than the instrumental route. The value of our integrative model is that it puts together the existing pathways in one model, with ideology as an extra pathway, and it provides a steering mechanism that explains which pathway to collective action will prevail. The research reported in this chapter extends our understanding of both collective action and regulatory focus. It documents the psychological importance of goals and self-regulatory processes as mechanisms affecting action taken in response to perceived goal obstruction and value violation. Importantly, it also demonstrates that regulatory focus analyses extend to group-level processes (see also Levine et al., 2000; Sassenberg et al., 2003; Shah et al., 2004). Individual differences in regulatory focus have implications for people’s motivation to engage in group-related action. The integration of two motivational mechanisms (regulatory focus and participation motives) in one model has proved useful in helping to explain who will participate in what social movement context, and why people are moved to engage in collective action. 140 Chapter 6. General discussion and conclusions. We started this dissertation with the observation that, so far, social psychologists have explored three possible motives that lead to participation in political protest: instrumentality, identity and group-based anger. Although supportive evidence abounds, many issues have been left unexplored (Klandermans, 2003). The first issue concerned the role of ideology. Another unanswered question was about the relative weight of the participation motives. Furthermore, Klandermans (2003) made a case for the study of identity in a more systematic way, and, lastly, he pointed to a set of unanswered questions concerning the dynamics between the individual protester and mobilization strategies of movement organizations. These unexplored issues guided our objectives in this dissertation. Our first objective was to integrate the three motives into a single model and to extend this theoretical model of political protest participation by introducing an ideology motive. Our second objective was to account for the influence of social movement context on the relative weight of the four participation motives. The third objective was to investigate whether regulatory fit explains why some collective action frames are more persuasive for an individual than others. Our final objective was to test whether regulatory focus functions as a steering mechanism that explains why one path to protest participation will prevail over another. We conducted two studies during two demonstrations to test our assumptions. In the following paragraphs we report the concluding remarks regarding these objectives. 6.1 An integrative model accounting for protest participation. The results of our studies provide considerable support for our integrative model accounting for the strength of the motives that inspire people to participate in political protest. Our studies replicated the findings of the dual path models of both Simon and colleagues (Simon et al., 1998; Stürmer & Simon, 2004; Stürmer et al., 2003) and van Zomeren et al. (2004). Importantly, though, our integrative model explained more variance than any earlier proposed model. Moreover, it appeared to be fruitful to extend the model by introducing an ideology motive, as people’s ideals and values generated passionate politics and explained considerable variance in their motivational strength to participate in political protest. Even in the more power-oriented protest staged by the trade union federations, ideological considerations played a major role in people’s decision to participate in such protest. Finally, our integrative model accounting for both CHAPTER 6 141 the direct and indirect effects of identity fitted the data better and explained more variance than a model merely accounting for the direct effects. We argued that it was unlikely that the four paths would function independent of each other and proposed that identification processes would have an integrating role. We hypothesized that group identification would influence not only why people participated, but also what they were angry about. Our results replicated the findings of Yzerbyt et al. (2003) in showing that the more the group is in them (i.e., the higher the group identification), the more protesters experience group-based anger. Collective action participation is seen as a way to show who we are and that we are angry. Interestingly, our study indicates that identification processes also influence the reasons for which people are angry. Indeed, the more “the group is in me”, the more group-based grievances or values are incorporated, and the more “I feel for us”. A strong identification with a social movement organization creates a sense of solidarity that spurs the motivation to take to the streets. Moreover, if strong identifiers feel that their group interests are threatened, they become angry and are even more willing to take to the streets. Weak identifiers, on the other hand, do not participate for reasons of solidarity or commitment but because of instrumental and/or ideological threats to their personal identity. Accordingly, weak instrumental and ideological considerations are accompanied by low levels of motivation, whereas strong instrumental and ideological considerations are accompanied by high levels of motivation. This seems to be in line with findings of Veenstra and Haslam (2000), who found that weak identifiers refrained from participation in union activities when reference was made to conflict alone, but that this effect was attenuated when reference was also made to threat. Weak identifiers seem to be more opportunistic in aligning with the group only when this serves to enhance their own personal identity (Ellemers et al., 1999). This post hoc explanation suggests that strong identifiers participate out of solidarity and shared grievances, whereas weak identifiers are more opportunistic. They participate because of a perceived threat to their personal identity, and alignment with the social movement organization seems to provide security in insecure times. The point of the matter, however, is that it suggests that future research of protest participation must take the degree of identification into account. Interestingly, a single model accounting for the combined motives of the two dual path models revealed that instrumental motives are completely translated into group-based anger. This seems in line with the findings of Stürmer and Simon (2004) that also show relatively little impact of the collective motive13. A closer look at their data reveals that participants placed a high value on the movement’s goals but considered the goals’ realization as not very likely. “This GENERAL DISCUSSION 142 lack of optimism [about the political efficacy of the social movement, JvS] may, at least partially, explain why the collective motive was ineffective” (Sturmer & Simon, 2004, p. 91). In our studies, though, the relatively small impact of instrumentality seems to be due to optimistic rather than pessimistic accounts of the perceived political efficacy of the protest event. Identification with a strong, effective, social movement organization, like the trade union federations with their ability to mobilize 250,000 protesters, enhanced strong feelings of group efficacy. Given these findings, one might wonder whether instrumental motives impact at all on people’s motivation to participate in collective action. Our findings suggest that identification with an effective and powerful organization, like the trade union federations, facilitates a transformation of the felt grievances into anger14. Malcolm X was aware of the power of emotions when he stated15 that “usually when people are sad, they don’t do anything. They just cry over their condition. But when they get angry, they bring about social change.” Indeed, feelings of solidarity and unity may give a sense of empowerment, which functions as the social glue that turns grievances into socially shared emotions like anger (Britt & Heise, 2000). Caution is advised in interpreting these findings, because these effects may be due to retrospection in the self-reported surveys. It may be that perceived political efficacy was high due to the fact that soon after the demonstration the government decided to withdraw the most “painful” cutbacks, e.g., early retirement rights. Yet, this still does not explain why political efficacy mediates effects between instrumentality and group-based anger for high rather than low identifiers. Apparently, identification with organizations staging power-oriented protest not only influences protesters’ motives and the reasons of their anger, but also communicates a sense of power. Our results illustrate that the logic of collective action involves much more than the instrumental pursuit of collective goods. In classic studies on social movements, instrumental movements were seen as movements that aimed at some external goal, whereas participation in expressive movements, on the other hand, was seen as a goal in itself. The distinction between instrumental and expressive movements fell into disuse, however, because it was thought that most movements had both instrumental and expressive aspects and that the emphasis on the two could change over time. Recently, the instrumental-expressive distinction has received renewed attention, this time to distinguish between the different motives people may have for participating in social movements. People may participate in social movements not so much, or not only, for 13 Their operationalization of a collective motive represents our operationalization of instrumentality. Mediation analyses revealed that for those who identified highly with the union (and not the alliance) political efficacy mediates between instrumentality and group-based anger (Sobel z-value = 2.14, p = .03). 15 Retrieved from the Internet, 2 April, 2005, http://www.quotationspage.com/quote/1879.html 14 CHAPTER 6 143 instrumental reasons but (also) to express their views and feelings (see Goodwin, Jasper & Polletta, 2001 for an overview). Our results are consistent with the view that, in addition to instrumentality and identification considerations, expressive motives like emotions and ideology create motivational energy for people to take to the streets. People are angry about some state of affairs or some government and develop feelings of moral indignation; they want to make that known and participation is seen as a strategy to solve their (collective) problems. They participate in a social movement not only to enforce political change, but also to identify with others involved, to regulate the emotions tied to the situation, and to maintain their moral integrity. Our results also illustrate that motivation to participate in collective action involves much more than conquering the free riders’ dilemma built into the instrumental pathway. It is true that in taking the instrumental route people have to overcome the free riders’ dilemma, but the remaining motives all generate an inner obligation that helps to overcome this dilemma—even though the three motives create a sense of inner obligation for different reasons. As mentioned before, group identification creates an inner obligation to behave as a “good” group member, ingroup norms are incorporated in the self, and this in turn results in an inner obligation to behave as a ‘‘good’’ group member. Ideology motives, in their turn, create a sense of moral obligation for reasons of moral integrity maintenance, whereas group-based anger, finally, creates a sense of inner obligation stemming from the emotional coping or catharsis function of collective action participation. The inner obligation attached to ideology and group-based anger becomes an energizing force that helps to overcome any free riders’ dilemma. The free rider literature tends to forget about these internal factors that push individuals toward participation. It tends to focus on the external pull factors, such as goal achievement and selective benefits. In our future research we want to put more emphasis on the push factors. 6.2 The influence of social movement context on the relative weight of the paths. Testing our model in different mobilizing contexts offered the opportunity to investigate the influence of mobilization context on people’s participation motives and the reasons for their anger. Our results seem to suggest that the campaigns of the different social movement organizations appeal to distinct individual motives, and reveal different reasons for people’s anger. As expected, in the context of the value-oriented protest only the ideology route influenced the motivation to participate, whereas the instrumental and, unexpectedly, the ideology route prevailed in the context of power-oriented protest. Few social movement scholars have actually linked the motivational configuration of individual protesters to the mobilization GENERAL DISCUSSION 144 context (for an exception see Klandermans, 1993), but our results suggest that it may be rewarding, using a comparative design, to study the influence of movement characteristics, such as type of action orientation, on the reasons for which people take part in collective action. Had we aggregated the data of both demonstrations and not taken into account the effects of context, we would not have discerned the different motivational patterns. Indeed, “without comparative studies of different campaigns, we would never be able to sort out these individual sources of variation” (Klandermans, 1993, p. 400). Moreover, aggregation of the data would not have revealed the interesting, but unexpected, finding of an ideology path in the context of power-oriented protest. Yet, on further consideration, the ideology path was not so unexpected in the context of trade unions. Various studies have found support for both an instrumental and an ideology route to union commitment and support, including studies examining motives for joining a union (De Witte, 1995), union commitment (Sverke & Sjöberg, 1997a) and trade union participation (Sverke, 1996). Indeed, “there is now general support for there being two main routes for union commitment and union support, the instrumental route and the ideological route” (Blackwood, Lafferty, Duck & Terry, 2003, p. 488). This has (at least) two potential implications for the results of our studies. It may indicate that, even in the context of power-oriented protest, ideological considerations have their influence on motivation to participate, or it may suggest that the action-orientation of the trade union federations was not “purely” power-oriented. The latter observation points to the fact that, according to Turner and Killian (1987), all three orientations play some role in every mobilization campaign. Future research might investigate whether the findings about instrumentality and ideology are related to this specific union social movement context or are a more general effect due to the combination of the four motives in a single model. 6.3 Regulatory organization-participant fit. Adding regulatory focus theory to our model enabled us to explain why people participate in one protest rather than the other. Our studies suggest that frame alignment may generate a regulatory organization-participant fit, that is, the more an organization emphasizes prevention aspects in its campaign, the more likely it is to attract prevention-focused protesters employing “typical” prevention motives, such as instrumental and identity motives; and the more it emphasizes promotion aspects, the more it tends to attract promotion-focused protesters employing “typical” promotion motives, such as ideology. The general purpose of this study was to consider how regulatory fit could influence the persuasiveness of a collective action frame. Our guiding CHAPTER 6 145 assumption was that persuasive messages like collective action frames usually involve some goal to be attained (e.g., moderation of general budget cuts proposed by the government) and some means described as the way to attain it (participation in collective action). This suggests that, when the arguments reflected in a persuasive communication, such as a call for action, fit the regulatory focus of the message recipient, the recipient should experience regulatory fit. This fit translates into a stronger persuasiveness of the message and, accordingly, people experiencing regulatory fit will be more inclined to comply with the call for action than people who do not experience such a fit. Our findings are consistent with the findings of Cesario, Grant and Higgins (2004) who show the importance of regulatory fit for persuasiveness of the message and agreement with its topic. However, the results of the former studies were acquired in a controlled laboratory setting, where the people read messages and had to indicate the extent to which the message was persuasive. Our studies suggest that, even in an uncontrolled field setting with an abundance of persuasive messages, the regulatory fit principle seems to operate. Our findings suggest that regulatory fit processes play a role in the way people construe their social and political world, and influence social and political decision making. This is consonant with findings of Camacho, Higgins and Luger (2003) who found that regulatory fit processes influence the evaluation of proposed policies. People had to evaluate proposals about a citywide policy change involving the New York public school system. These proposals were stated either in eager terms (“The primary reason for supporting this program is because it will advance children’s education and support more children to succeed”, Camacho et al., p. 507) or in vigilance terms (“The primary reason for supporting this program is because it will secure children’s education and prevent more children from failing”, Camacho et al., 2003). Promotionfocused individuals judged the proposals defined in eager terms as more morally right than the vigilant programme, whereas prevention-focused individuals judged the vigilant programme as more morally right than the eager programme. As mentioned previously, our data do not allow us to disentangle the (interrelated) regulatory effects of group processes such as recruitment, group identification, group polarization and collective action frames relating to protest participation. However, when an individual participates in collective political action staged by a social movement organization, this is the result of a sometimes lengthy process of mobilization. Successful mobilization gradually brings what Klandermans (2003) calls demand and supply together. If substantial proportions of the population are aggrieved, and if movement organizations stage collective action to voice those grievances, a massive protest movement may develop. Our theoretical framework is a first cautious step in studying the complex relation between demand and supply; this may make it a GENERAL DISCUSSION 146 fruitful bridge builder between the micro, meso and macro levels of collective action, but it also leaves a lot of questions regarding the connection between demand and supply unanswered. The finding of an organization-participant regulatory fit may have practical implications for mobilization attempts of organizations. Effective communication in both the stadium of consensus and of action mobilization is of vital importance to social movement organizations. Regulatory fit may make messages more appealing—even more so because regulatory focus can be temporarily induced and this may enhance the resonance of a message. 6.4 Regulatory focus as steering mechanism. Our last objective was to incorporate a motivational steering mechanism into our theory in order to explain why people are inclined to take one route rather than another to protest participation. Indeed, with this last objective we hoped to determine whether regulatory focus controls what path people will take to protest participation. The basic assumption steering this objective was: if regulatory focus can be seen as a general worldview and has cognitive, emotional and behavioural consequences for how individuals respond to the world (Higgins, 1998), it may also influence a frame to interpret the social and political world, and therefore shape protesters’ participation motives. Our studies seem to suggest that the paths taken to protest are indeed moderated by individual, idiosyncratic regulatory processes, although the effects are moderate. Prevention-focused people are inclined to take the instrumental and/or identity route in prevention-oriented protest, whereas promotion-focused people are inclined to take the ideology route both in prevention- and promotion-oriented protest. Our findings imply that people’s specific political motives arise from particular attitudes or values and their general worldview. Regulatory focus explained some variance in the motives employed (10%), but definitely not all. Indeed, regulatory focus explained to a certain extent why people are inclined to take one route rather than another, but there is certainly room for improvement. However, multiple statistical methods revealed the same results over and again, and the fact that, in all studies, the pattern was in the expected directions suggests some robustness. This makes us confident that incorporating regulatory focus as a motivational component in our protest participation theory is a first step in the right direction. Indeed, as indicated, there is room for improvement, so future research could investigate other (social) psychological motivational components that could explain why people take one route rather than another. Importantly, though, our studies also demonstrate that regulatory focus analyses extend to group-level processes (see also Levine et al., 2000; Sassenberg et al., 2003; Shah et al., 2004). CHAPTER 6 147 Individual differences in regulatory focus have implications for people’s motivation to engage in group-related action. Participating in one protest rather than another may be such a group-related action. As mentioned before, compared to the union protest, the Turn the Tide protest attracted relatively many non-members (19% and 44%, respectively). New social movements like the alliance have “informal modes of belonging, including shared concern about diverse issues and identity politics. People see themselves as belonging simply by ‘turning up’ or sharing political sympathies with an easy-entrance” (Norris, 2004, p. 10.). These looser boundaries and informal modes of belonging may be related to the fact that Turn the Tide attracted relatively more promotion-focused people. Promotion-focused people prioritize selfdetermination and autonomous values; participating in a protest event as a non-member might be seen as such an autonomous independent act that fits their dominant focus. From these results, two main issues emerge which, in our opinion, deserve further discussion, namely, protest participation and identity, and protest participation and self-regulatory strategies. 6.5 Protest participation and identity. We assigned identification with a social movement organization a central role in our model explaining protest participation. In doing so, we refined and enlarged the understanding of identification processes in protest participation. We replicated the finding that identity motives feed directly into the strength of motivation, indicating that identification with others involved influences the strength of motivation sufficiently to take to the streets. Importantly, though, our results suggest that identification played a vital role in the other path taken to protest participation as well. The more strongly people identified with one of the two movement organizations, the more group-based grievances or values were incorporated and in turn more group-based anger was experienced. Identification with the group at stake influences what people want to accomplish by actually participating in protest activities and therefore creates shared group-relevant motives. These contextually related differences in the valence of people’s feelings, thoughts and participation motives are a reminder that collective action involves an individual thinking and acting in concert with others. Collective action is contingent on seeing the self as part of a group. Indeed, acting collectively requires some collective identity or consciousness (Klandermans & De Weerd, 2001). Our findings are important in confirming what one would expect—the level of attachment people have to social movement organizations influences the attitudes, feelings and behaviour GENERAL DISCUSSION 148 relating to this group. The stronger someone’s group identification, the more shared beliefs, grievances and fate comprised in the group’s collective identity are incorporated in the individual’s collective identity. At the heart of every social movement organization is the desire to change society. Indeed, social movements do their utmost to explain how they interpret a social, political or economic change (its diagnosis) and what should be done (prognosis) as a reaction to perceived losses or unfulfilled aspirations. In communicating its prognosis, the social movement organization will emphasize its own role in achieving the goal. At the psychological level, then, a social movement organization may provide meaning and it may facilitate individuals to cope with their social and political environment. Indeed, the results of our studies indicate that, while identification impacts upon the strength of the motivation to participate in collective action, identification also influences people’s coping strategies. In the power-oriented protest, identification with the social movement organization reinforced instrumental motives, whereas, in the value-oriented protest, identification reinforced ideology motives. In other words, identification with an organization staging power-oriented protest feeds more into the problemfocused coping route (i.e., instrumental path) to protest participation, whereas identification with an organization staging value-oriented protest feeds into the emotional coping route. Identification with an organization that emphasizes its effectiveness and power in the political arena increases instrumental motives, because this identification enhances group efficacy and instrumental social support. Identification with an organization that emphasizes the ideology behind its goals (i.e., value-oriented) increases ideology motives, because this identification helps people to perceive an illegitimate and unfair situation as a collective disadvantage. Hence, information about the social support for one’s own opinions may help to define the experience as collective and shared and the situation as group-based. Accordingly, the function the group fulfils is different: on the one hand, a shared power function: “together we are strong” and, on the other hand, a shared reality function: “together we know”. One would expect people to be angrier at the demonstration (i.e., in the face-to-face interviews) than at home (i.e., in the questionnaires). However, contrary to what one would expect, the union protesters are angrier at home, whereas the Turn the Tide protesters are angrier during the demonstration. This may result from the different functions of social movement organizations (“together we are strong” and “together we know”) and the effect of the mere presence (e.g., Zajonc, 1966) of ingroup members on people’s feelings. In the context of the protest staged by the unions, the “together we are strong” function prevails and, accordingly, the mere presence of many other ingroup members may give the impression that the goal will be CHAPTER 6 149 reached. The impression that the goal will be reached may enhance feelings of relief and reduce the anger. In the context of value-oriented protest, however, the “together we know” function prevails, and people participate because they want to express their view. The mere presence of ingroup members sharing the same view may increase the anger about an ambiguous motive such as ideology. “For those in the midst of the collectivity, their entire sensuous universe (sights, sounds, even smells) confirms their worldview” (Reicher, 2005). The group protesters as observable realistic presence may function as a confirmation of one’s ideological view and may enhance anger. This suggests that the mere presence of other ingroup members impacts differently on emotions experienced, contingent on identification and social movement context. Yet, these are just ad hoc speculations on an unexpected finding. Future research might study the effects of the mere presence of ingroup members on motives and emotions as a result of social movement context and identification. 6.6 Protest participation and self-regulatory strategies. Mental models of why people engage in protest participation seem to be shaped by the mental models they have about the social world. Whether one has a mental model that tends towards avoiding something (i.e., prevention-focused) or a mental model that tends towards attaining something (i.e., promotion-focused) has profound implications for emotions, thoughts and behaviours. When one encounters adverse environmental, social, political or economic changes, an imbalance between what one needs and what one actually encounters as a result of this adversity can motivate actions directed toward preventing or promoting this change. One possible action can be participating in (or even organizing) social action. Besides distinct goals, needs and values, the regulatory foci have distinct motivational “energy”. Since we interviewed people who were already participating, all our respondents, obviously, had enough motivational energy to take to the streets. However, to test our model in a more dynamic research design that would predict who would participate, an elaboration of this motivational energy might be useful. Promotion-focused people are concerned with the presence or absence of positive outcomes and their strategic inclination is approach in a state of eagerness, whereas prevention focused people are concerned with the presence or absence of negative outcomes and their strategic inclination is avoidance in a state of vigilance (Higgins, 1997; 1998). Recent findings suggest, however, that strategic inclinations may differ as a function of threat. Stroesnner and colleagues (2005), for instance, theorized that a prevention focus combined with negative or threat-related information in the environment would change the GENERAL DISCUSSION 150 typical information-processing strategies associated with the regulatory state. Maintaining security when there is a threat in the environment requires an active rather than passive form of vigilant behaviour. Pham and Higgins (2005, p. 34) propose that “if the current state is highly undesirable, prevention-focused individuals will be more likely to pursue ‘riskier’ options that could remove the undesirable state—thereby, exhibiting greater ‘risk-seeking’ behaviour—than promotionfocused individuals.” This is because prevention-focused individuals would consider it a mistake to remain in the current state, and feel it necessary to choose the “riskier” option. Note that such seemingly “risk-seeking” choices under prevention would arise not really because preventionfocused individuals want to take risks, but rather because they see it as a necessity. This suggests that prevention-focused people will only choose to participate if they see the feasibility of improvement. One may, of course, wonder why we chose regulatory focus theory above one of the many other motivational theories. First, and most important, regulatory focus theory merged gently into our theory because it could help to explain why people with distinct goals, needs and values take different routes to protest. Indeed, protest participation is often activated by obstruction of goals and needs and violation of values. Moreover, many motivational theories include approach and avoidance as core themes (aversive and appetitive motive systems, e.g., Carver & White, 1994; Harmon-Jones & Allen, 1997; BIS/BAS (behavioural inhibition system, behavioural approach system, e.g., Gray, 1972; 1994a; approach-avoidance motivation, e.g. Elliot & Thrash, 2002). All these motivational theories hold in common that an avoidance system is activated by the perception of threat, the result is inhibition or withdrawal behaviour, whereas when given cues of incentives people engage in more approach behaviour. Higgins’ regulatory focus theory is a model of approach and avoidance that echoes the core themes of approach and avoidance. Yet, regulatory focus theory has more to offer, because not only does it account for approach and avoidance motivation, but it also helps to explain what people want to approach and avoid. Moreover, regulatory focus theory appears to be helpful in explaining a regulatory organization-participant fit. To the best of our knowledge, other motivational theories do not account for this kind of fit mechanism. Van Zomeren et al. (2004) assumed coping mechanisms as underlying psychological constructs explaining why people are willing to participate in protest, whereas we integrated selfregulatory mechanisms. Coping consists of activities undertaken to master, reduce or tolerate environmental demands perceived as representing potential threat, existing harm or loss (Lazarus CHAPTER 6 151 & Folkman, 1984). Self-regulation is defined as the process through which people control, direct and correct their own actions as they move toward or away from various goals, [needs or values, JvS] (Carver, 2001). Although the literatures on coping and self-regulation have developed largely in isolation, they share a fundamental concern with the relation of personal, social and situational factors to people’s emotions, thoughts and behaviours as they anticipate or encounter adversity (Aspinwall, 2004). Our results show that whether one is coping to attain something or to avoid something has profound implications for the participation motives employed and the reasons for one’s anger. A last remark on regulatory strategies relates to goal processing. One might wonder whether taking one route rather than another to protest participation can be equated with conscious reasoned action or whether it is a result of an unconscious reflex like a knee jerk reflex. We assume it would be misleading to consider them as strategies deliberately chosen to take one route rather than another. Yet, we do not perceive of the process as completely unconscious either. Talking to protesters during demonstrations and asking them the question: Why are you participating in this protest? reveals that people can explain why, and that they show the ability to be reflective on decisions like this. 6.7 Possible shortcomings. Prior to discussing the implications of our studies for future research, we must recognize the limitations of the current studies. Perhaps no area in mass political behaviour research is plagued with such serious methodological difficulties as the study of participation in political protest and other unconventional activities. A fundamental problem is that, unlike voting for example, protest participation does not usually occur during fixed periods, and hence researchers typically cannot plan a well-thought-out study in advance. Moreover, data collection on protest participation is rarely comprehensive, and scholars have to come up with creative solutions to avoid problems of reliability and validity. Due to these two challenges, the lion’s share of our shortcomings relates to methodology. Our design called for us to look at those people who were actually participating in one of the two protest events. This implies that we sampled on the dependent variable. Indeed, this research design does not permit us to state who will participate and who will not. However, it enables us to draw conclusions on the motivational patterns as a function of the social movement context. Moreover, our dependent variable—motivational strength—showed enough variability GENERAL DISCUSSION 152 to enable us to study the variability of the motivational concepts as a function of variability in motivational strength. Despite the comparative design, it remains difficult to arrive at robust conclusions about the causal link between the concepts included in our integrative model. The comparisons that we made of social movement context are useful, but field studies do not have the rigour needed for solid tests of causality. Manipulating one of the concepts, for example identification or regulatory focus, would offer the opportunity to observe whether the paths taken to protest participation—or intentions to participate—change as a consequence of the manipulation. It would be worthwhile trying to replicate our findings in a more controlled laboratory setting. Another shortcoming, not relating to methodology but rather to operationalization, is about the operationalization of ideology. In our studies, ideology is operationalized using two emotion items (“I am protesting because: I am worried about the proposed government policy”, “I am concerned about the proposed government policy”); two injustice items (“I find the proposed government policy unfair”, “I find the proposed government policy unjust”); one accountability item (“I want to take my responsibility”); and one “against my values” item (“The proposed government policy is against my principles”). Although we are reasonably confident that we captured the ideology motive with these items, there is room for improvement. Currently we are working on a measure that more closely represents the conceptualization of ideology (an event in the social, political or economic environment is against values, people want to express their view, participation for reasons of moral integrity maintenance, and participation in political protest functions as an emotional valve). It is important to acknowledge other limitations of our model. We do not claim that the proposed model is complete, because we have integrated only some elements of different approaches. More research is needed to enrich our model. Moreover, despite the fact that we are reasonably confident in the proposed model because it is consistent with previous formulations and we were able to rule out a few alternative models, by no means have we ruled out all plausible alternatives, and researchers are encouraged to consider alternative formulations. 6.8 Protest participation and directions for future research. 6.8.1 From a static to a dynamic model. One of the most interesting contributions of this dissertation in our view is the integrative model accounting for motivational strength in protest participation. We asked ourselves in the introduction to this dissertation: “Why would someone decide to go to Amsterdam and CHAPTER 6 153 participate in a mass demonstration?” Our integrative model helps to provide an answer to this question. However, many questions still remain unanswered. Of course, the first, and probably most important, question is the extent to which our integrative model enables us to predict protest participation. Given the fact that the building blocks (the dual path models) of our integrative model are used to predict intentions to participate, it is expected that our integrative model should predict protest participation too. However, in the above, we have already mentioned that, to test for that, protesters should be interviewed before and after the event. We suggest taking that one step further and moving from a static to a more dynamic model in an (Internet) panel data framework. A more dynamic approach would provide the opportunity to study concepts like identification, participation motives, empowerment, emotions and political efficacy as a consequence and an antecedent of collective action. “From an investigational point of view, it is difficult to deal with a variable that, at the same time, can be a dependent and an independent variable, can develop over time or change across contexts” (Ellemers et al., 1999, p. 3). Yet, studying protest participation in a more dynamic way would do more justice to the theoretical and empirical richness of the concepts and may be crucial to gain better insights into the processes at hand. 6.8.2 Effects of regulatory focus on protest participation. The studies reported here investigated the extent to which self-regulatory strategies control what path people take to protest participation. Yet, as mentioned previously, our research setting allowed us to test some hypotheses of our integrative model, though not all. Our research setting was the often-investigated politically leftist oriented protest of the Turn the Tide alliance and the labour movement. The first protest was aimed at typical, progressive liberal values and, according to our model, would attract promotion-focused people, whereas the latter, aimed at social security, would attract prevention-focused people. In the case of less common movements or protest events, however, one might encounter completely different patterns. Take, for example, the conservative pro-life movement. In this case, our model would predict that pro-choice policies (abortion rights) might be perceived as a violation of conservative and traditional values for prevention-focused people; this may evoke motivation to participate in pro-life movements with an ideology motive. Another example would be a student’s movement staging a protest event aimed at reducing tuition fees. In terms of regulatory focus, increasing tuition fees might be seen as an obstruction of psychological growth and development, or achievement needs; as promotion-focused persons are motivated by achievement needs, we expect them to employ an GENERAL DISCUSSION 154 instrumental motive in protests with these kinds of claims16. Testing these propositions would be an interesting avenue for future research. In Chapter 5 we proposed a set of assumptions regarding the influence of regulatory mechanisms on processes like recruitment, group identification, group polarization and the definition of collective action frames. Our data enabled us to assess the fit between the regulatory characteristics of a collective action frame and the regulatory focus of individual protesters. This implies that a set of assumptions relating to the dynamic interplay of regulatory characteristics between “demand and supply” remains untested. First, we assumed that prevention-focused individuals are attracted by social movement organizations with missions, goals and ideologies stated in prevention rather than promotion terms, and that promotion-focused individuals are attracted by social movement organizations with missions, goals and ideologies stated in promotion rather than prevention terms. Next, we made assumption about group polarization in times of intergroup conflict. More specifically, we argued that prevention characteristics of prevention-oriented social movement organizations would intensify, whereas promotion characteristics of promotion-oriented social movement organizations would intensify. Since our results suggest a fit between the regulatory characteristics of calls for action and the regulatory characteristics of individual protesters, investigating the above-mentioned assumptions regarding the dynamic interplay of regulatory mechanisms of intra- and intergroup processes and individual regulatory focus would be worthwhile. 6.8.3 Effects of a variety of emotions on protest participation. Our results suggest that (nearly) all the routes to protest participation are accompanied by groupbased anger. For those of us who have been part of protest events or watched reports on protest events in the news media, these findings are hardly surprising. Indeed, it is hard to conceive of protest detached from emotions. In the current model, we have considered group-based anger as the relevant action-based emotion fostering collective action, but other emotions may be relevant in stimulating protest participation. Therefore, not only group-based anger but a variety of groupbased emotions may stimulate protest participation. Indignation is one of the emotions that might stimulate protest participation, because it “puts fire in the belly and iron in the soul” (Gamson, 1992, p. 32). Indeed, indignation is anger about the mistreatment of someone or something (Ekman, 1993); this makes indignation the most 16 Our data may give some indication in that direction. As mentioned earlier, the general budget cuts of the government were comprehensive, and the two movements emphasized different aspects of the proposed policies, the alliance emphasized (among other things): “College fees are under fire as well”. An instrumental motive aimed CHAPTER 6 155 political of feelings (Reichenbach, 2000). Therefore, it may be interesting to investigate the role of indignation in stimulating protest behaviour (cf. Jasper, 1997; 1998; Goodwin et al., 2001; Kim, 2002). Critical readers familiar with regulatory focus theory may well ask why we did not take the various emotions related to the regulatory characteristics into account. And they are right; our studies do not answer questions concerning the emotional dimensions of the regulatory foci. There is evidence, however, that obstruction of an ideal goal feels different from the obstruction of an ought goal. Obstruction of an ideal goal evokes dejection-related emotions, whereas obstruction of an ought goal produces agitation-related emotions (see for example Higgins, 1998; Higgins et al., 1997; Higgins & Brockner, 2001). Therefore, it might be worthwhile to study whether emotions stimulating protest participation differ when protest participation stems from obstruction of an ought goal as distinct from obstruction of an ideal goal, particularly as Bizman, Yinon and Krotman (2001) show that the emotional dimensions explained by regulatory focus theory could be extended from individual-based to group-based emotions. Moreover, it is possible that whether one could or could not do something about goal obstructions may make a difference for the emotion experienced. Research shows that emotions result from the appraisal of the implications of an event for goals of the individual and his or her ability to cope with the consequences of the event (Scherer & Zentner, 2001). Therefore, perceived political efficacy may be viewed as one of the possible operationalizations of coping potential, with strong political efficacy eliciting anger, and weak political efficacy eliciting sadness (cf. Scherer & Zentner, 2001). Hence, conceptualized in this way, emotions would not function as a separate route to protest participation, yet would be dependent on the constellation of the regulatory focus and perceived political efficacy. 6.8.4 Effects of political, economic and cultural contexts on protest participation. In the studies reported here, a comparison across protest events helped to explain how contextual variation influenced the relative weight of various paths. However, in classic survey-based studies on protest participation, political protest was understood as arising from an interaction between individual characteristics and collective actors, such as movement organizations. Moreover, this interaction seemed to be shaped by different political and socio-economic conditions of countries, such as the duration of democracy and economic wealth (Barnes & Kaase, 1979). Recently, it has been widely acknowledged that the dynamics of participation are at college fees is more strongly related to promotion-focus than prevention-focus. This suggests that promotionfocused protesters can take the instrumental pathway aimed at changing a threat to their achievement goals. GENERAL DISCUSSION 156 created and limited by characteristics of the national contexts in which people are embedded (see Roggeband, 2002). The influence of national context on protest participation is determined by the political, socio-economic (Barnes & Kaase, 1979; Roller & Wessels, 1996) and cultural (Polletta, 1999; Goodwin & Jasper, 1999) characteristics in which people live. The political, socio-economic and cultural characteristics influence and shape a mental model about what the social world looks like and what it ideally should look like. Indeed, these collective mental models may create and limit goals, aims, objectives opportunities (Polletta, 1999) and (group-based) emotions (Goodwin et al., 2001) of both individuals and organizations, and therefore may shape the reasons why people participate in protest. Yet, different countries have different collective mental models that impact on the individual and organizational mental models. The results of our studies seem to suggest that an interaction of organizational attachment and individual differences influences why people engage in protest participation. However, in our studies people were living in the same national context. Therefore, it would be interesting to investigate whether various national contexts, that is, various collective mental models, impact differently on the relative weight of the paths to protest participation. In a comparison of various national contexts, several concepts may be relevant. One might think of differences in social capital, specifically the “connections among individuals, social networks and the norms of reciprocity and trustworthiness that arise from them” (Putnam, 2000, p. 19). The level of social capital of a country is believed to facilitate political participation. Likewise, there might be differences in dominant regulatory characteristics of a country. When a country has a collectivistic mental model, people will tend towards avoiding something (i.e., they will be prevention-focused), whereas with an individualistic mental model people will tend towards attaining something (i.e., they will be promotion-focused, Lee, Aaker & Gardner, 2000). It may be that the regulatory characteristics of a country sensitize a discrepancy between what people actually encounter and either what people ought to have or what people ideally want to have. As a result, this discrepancy may motivate actions directed toward preventing or promoting social change; and, although we can conceive of many other relevant concepts, we will restrict ourselves to level of modernity as the last concept. Roller & Wessels (1996) show that increasing levels of modernity are accompanied by increasing levels of protest participation Some researchers have argued that the general prosperity of advanced industrial societies has heightened the relative importance of cultural issues for an increasing number of people who are not preoccupied with economic survival; this, they say, has led to a sizeable shift in advanced industrialized nations toward postmaterialist values, and these values find expression within new social movements (Inglehart, 1987). This may indicate that the higher the CHAPTER 6 157 level of modernity, the more people are inclined to take the ideology route, whereas a lower level of modernity may lead to more instrumental concerns underlying protest participation. These findings suggest that it would be interesting to investigate these various collective mental models and their influence on organizational and individual mental models and motivational dynamics towards protest participation. Comparative research of the same movement in different national contexts may be a key venue to investigate the moderating effects of national context. In our research paradigm, our comparison of the motivational dynamics to participate in the more traditional trade union federations and the new social movement Turn the Tide was made in the same national context. It would be interesting to extend this paradigm by a comparison of the same movements, for example the trade union federations and the anti-globalization movement, in different national contexts. Such comparisons are important because they may reveal diverging political, economic, cultural or social psychological dynamics of movement participation. 6.9 Concluding remarks. Based on previous protest literature, we proposed and tested an integrative model accounting for protest participation and extended it by introducing regulatory processes controlling which paths are taken to protest participation. Promotion-focused protesters are inclined to take the ideology route to both promotion- and prevention-oriented protest, whereas prevention-focused protesters are inclined to take the instrumental/identity route to prevention-oriented protest. These results add to our knowledge about protest participation in many ways and at the same time raise new and interesting questions for future research. In reaction to irrational and emotional explanations, the dominant academic political analyses on protest participation have shifted to rationalistic, structural and organizational explanations. Our research, however, suggests that, by reducing protest participation to rational, structural and organizational processes, researchers appear to have swung the pendulum too far in the opposite direction. Contingent on an interaction of individual and organizational characteristics, protesters participate on the basis of instrumental, solidarity and expressive considerations. From a rational man point of view, all groupmembers would choose for the freeriders option. However, reality shows that people time and time again even if death follows are prepared to participate in political protest. This seems to suggest, that future protest studies could improve by integrating the more rational, structural and organizational explanations with the more irrational and emotional explanations. GENERAL DISCUSSION 158 159 Samenvatting In dit proefschrift wordt getracht een antwoord te geven op de vraag waarom mensen meedoen aan politiek protest. Politiek protest is een ‘normaal’ onderdeel geworden van het politieke proces in Westerse democratische samenlevingen en bovendien doen steeds ‘normalere’ mensen mee aan politiek protest (Norris, Walgrave, & van Aelst, 2005). Een groepslid neemt deel aan politiek protest elke keer als hij of zij zich gedraagt als afgevaardigde van die groep en als de actie gericht is op het verbeteren van de condities van de groep als geheel (Wright, Taylor & Moghaddam, 1990). De laatste twee decennia zijn sociaal psychologen begonnen met het onderzoeken van individuele participatie in protest. Eerst toonden zij aan dat participatie in protest werd gecontroleerd door instrumentele overwegingen (Klandermans, 1984), dat is als protest participatie wordt gezien als een mogelijkheid om een een (collectief) probleem op te lossen tegen aanvaarbare kosten. Geleidelijk werden meer en meer motieven onderzocht die mensen stimuleerden deel te nemen aan protest. Het eerste motief dat werd toegevoegd, was identificatie (Simon et al., 1998), dat wil zeggen, mensen participeren omdat ze zich identificeren met anderen die zich ook inzetten voor de strijd. Onlangs is daar ‘groeps-gebaseerde woede’ bijgekomen (van Zomeren et al., 2004), hier is deelname aan protest een uitlaatklep van de emoties die een bepaalde sociale of politieke misstand oproept. Zowel Simon en collega’s en van Zomeren en collega’s stellen een dual-path model voor, Simon en collega’s laten de instrumentele en identiteitsroute zien en van Zomeren en collega’s onderzochten de instrumentele en de groepsgebaseerde woede route. Het eerste doel van dit proefschrift was het intergreren van deze routes in één model, dus een instrumentele, identiteits en groeps-gebaseerde woede route naar protest. Wij misten echter een in onze ogen belangrijk fundamenteel motief tot actie: ideologie. Mensen die de ideologische route nemen, participeren omdat een stand van zaken tegen hun normen en waarde syteem indruist. Het tweede doel van dit proefschrift was dan ook het uitbreiden van het theoretisch model door de toevoeging van een ideologie route. Aldus werd in dit proefschrift een model voor protest participatie met vier motieven getoetst: instrumentaliteit, identiteit, ideologie en groepsgebaseerde woede. Zowel Simon en collega’s alswel van Zomeren en collega’s laten zien dat de instrumentele, identiteits en groep-gebaseerde woede route onafhankelijke paden naar protest zijn en wij waren in de veronderstelling dat dat ook voor ideologie geldt. Echter, de groepsgebaseerde emotie literatuur stelt dat naarmate de identificatie met een groep toeneemt, de groeps-gebaseerde emoties sterker zullen zijn (Yzerbijt et al., 2003). Wij hebben dit SAMENVATTING 160 doorgetrokken door te veronderstellen dat identificatie niet alleen beïnvloedt dat mensen woedend zijn maar ook waar ze woedend over zijn. Dit impliceert dat de participatie routes niet onafhankelijk kunnen zijn. Het derde doel van dit proefschrift was dan ook het incorporeren van zowel directe als indirecte routes in het participatie model. Dit participatiemodel begint dan met identificatie en loopt, afhankelijk van de mobiliserende context, ofwel via instrumentaliteit ofwel via ideologie naar groeps-gebaseerde woede. Identificatie heeft dus een prominente plaats in het model gekregen. Voor de mobiliserende context baseerden wij ons op het concept actie orientatie van Turner en Killian (1987). Deze auteurs delen protest acties in als (a) power-georiënteerd, (b) participatie-georiënteerd, en (c) value-georiënteerd. Power-georiënteerd wil zeggen gericht op het verkrijgen en uitoefenen van invloed, participatie-georiënteerd is gericht op de voordelen van participatie en value georiënteerd is gericht op de doelen en de ideologie van de beweging. Identificatie met een power-georienteerde organisatie betekent dat motivatie om te participeren voortkomt uit woede over instrumentele aspecten en bij een value-georienteerde organisatie uit woede over ideologische aspecten. De huidige participatie theorieën van Simon en collega’s en van Zomeren en collega’s besteden geen aandacht aan welk motief voor wie prevaleert en waarom? Dergelijke vragen veronderstellen een ‘besturingsmechanisme’. Dit brengt ons bij het vierde doel van dit proefschrift: wij voegen een besturingsmechanisme aan ons participatie model toe dat kan uitleggen waarom op een bepaald moment de ene route de voorkeur geniet boven de andere route. Wij stellen regulatieve focus voor als besturingsmechanisme. Regulatory focus theory is een motivatietheorie gebaseerd op zelf-regulatie (Higgins, 1997; 1998). Deze zelf-regulatie mechanismen kunnen promotie-focused zijn, bij het nastreven van behoefte aan koestering, ideale doelen (hoop en wensen), en waarden als self-directed and autonomie of prevention-focused, bij het nastreven van veiligheidsbehoeften, ‘ought goals’ (verplichtingen en verantwoordelijkheden), en waarden als traditie en conformiteit. Dus, de twee foci komen voort uit verschillende behoeften, doelen en waarden en verschaffen daardoor verschillende antwoorden op de vraag “Wat is mijn relatie tot de wereld?” Strauman (1996) refereert aan deze vraag door regulatieve focus te beschrijven als een worldview—een geneigdheid om situaties en ervaringen te beschrijven in termen van de twee foci. Regulatieve focus kan dus gezien worden als een algemeen wereldbeeld en heeft cognitieve, emotionele en gedragsmatige gevolgen voor hoe mensen reageren op hun omgeving (Higgins, 1998). Wij stellen dat regulatieve focus beïnvloedt hoe mensen hun sociale en politieke omgeving interpreteren en daarmee vorm geeft aan de motivationele constelatie van demonstranten. SAMENVATTING 161 Ons vijfde en laatste doel was te onderzoeken of regulatieve focus beïnvloedt welk collective action frame voor wie overtuigend is. Met andere woorden, trekken organisaties die hun protest activiteiten in promotie termen framen meer promotie- dan preventie-focused demonstranten aan en trekken organisaties die hun protest activiteiten in preventie termen framen meer preventie- dan promotie-focused demonstranten aan? Dergelijke vragen hebben betrekking op de dynamiek tussen individuele demonstranten en mobilisatie strategieën van sociale bewegingen, met andere woorden persoons-omgevingsfit, of in Higgins’ (1997; 1998) woorden, regulatory fit. De relatie tussen participatie motieven (instrumentaliteit, identiteit, ideologie en groepsgebaseerde woede), zelf-regulatie en mobiliserende context vormen samen de rode draad door dit proefschrift. Bestaande instrumenten om chronische regulatieve focus vast te stellen waren niet geschikt om toe te passen in een veldstudie met ‘gewone’ mensen in tegenstelling tot studenten. Alle voorhanden zijnde instrumenten vragen de participant te reflecteren op zijn of haar jeugd, dit is geschikt om aan studenten voor te leggen, echter een vakbondslid van 78 zal op zijn minst verbaasd zijn dergelijke vragen te moeten beantwoorden. Andere voorhanden zijnde instrumenten maakten gebruik van computers, dus ook deze konden in ons veldonderzoek niet gebruikt worden. Dit deed ons doen besluiten om zelf een instrument te ontwerpen wat breed toepasbaar is. Het werd een instrument gebaseerd op spreekwoorden, met bijvoorbeeld “wie niet waagt, wie niet wint” als promotie-item en “voorkomen is beter dan genezen” als preventie-item. In Hoofdstuk 2 van dit proefschrift worden 3 studies beschreven waarin het instrument (RFQproverb) wordt gevalideerd. In de eerste studie werd gekeken naar de voorspellende waarde door RFQ-proverb te relateren aan studie motivatie. In een paper-pencil test werd een aantal vragen over studie motivatie en RFQ-proverb voorgelegd. Wij voorspelden en vonden dat intrinsieke studiemotivatie samenhing met promotie-focus en dat extrinsieke studie-motivatie samenhing met preventie-focus. Om deze resultaten te repliceren en om RFQ-proverb te relateren aan constructen die, theoretisch, zouden moeten correleren met de RFQ-proverb subschalen (promotie en preventie) werd een tweede studie opgezet. Ook werd een bestaande RFQ meegenomen. Naast dat de resultaten uit Studie 1 werden gerepliceerd, liet deze studie zien dat de subschalen van RFQ-proverb zinvol correleerden met de subschalen van de andere instrumenten. De laatste valideringsstudie was een Internetstudie die ons de kans bood om te kijken naar de toegankelijkheid van het promotie- en preventie-construct. Volgens Fazio (1995) is de responsetijd een weergave van hoe vaak een construct wordt gebruikt, dus mensen met een chronische promotiefocus zouden sneller antwoord moeten geven op de promotie- dan de SAMENVATTING 162 preventiespreekwoorden terwijl mensen met een chronische preventie-focus sneller antwoord zouden moeten geven op preventie- dan promotiespreekwoorden. Dit was precies wat we vonden. De 3 studies laten zien dat RFQ-proverb een instrument is met goede voorspellende waarde en schaalkarakteristieken. Bovendien bleek RFQ-proverb sterk gerelateerd aan de reeds bestaande RFQ en hing het op de verwachte manier samen met instrumenten als Need for Cognition e.d.. Naar aanleiding van de 3 valideringsstudies concludeerden wij dat RFQ-proverb een valide en betrouwbaar instrument is. Bovendien is het een discrete meting die toepasbaar is in vele settings waar regulatieve focus het te onderzoeken concept is. Dit deed ons besluiten om RFQ-proverb toe te passen in onze studies naar protest participatie. Op 2 oktober 2004 vonden in Amsterdam tegelijkertijd twee grote demonstraties plaats. Een demonstratie op het Museumplein was georganiseerd door de vakbeweging en een andere demonstratie op de Dam was georganiseerd door Keer het Tij (een platvorm van 550 organisaties met meer dan 500.000 leden verenigd tegen het neo-liberale beleid). Tijdens deze demonstraties hebben wij data verzameld om ons participatie model te toetsen. Daarnaast bood deze comparatieve setting een unieke kans om de invloed van mobiliserende context op motivationele constellatie van individuele demonstranten te onderzoeken. In onze onderzoeken naar protest participatie hebben wij gekozen voor de protest survey method van Walgrave en collega’s (van Aelst & Walgrave, 2001). Deze methode kenmerkt zich door mensen te ondervragen tijdens demonstraties, dus in het heetst van de strijd! Echter, wetenschappelijk onderzoek kenmerkt zich door controleerbaarheid en repliceerbaarheid, hoe is het mogelijk betrouwbare en valide data in een kolkende massa mensen te verzamelen? Walgrave en collega’s hebben twee technieken ontworpen die dit mogelijk maken. De eerste techniek is ontworpen om (mogelijke) problemen van non-response te ondervangen. Hiertoe werden allereerst korte face-to-face interviews over een aantal demografische variabelen en de belangrijkste onafhankelijke variabelen afgenomen waarna post vragenlijsten werden uitgedeeld die demonstranten thuis in konden vullen en opsturen. Omdat (bijna) niemand een face-to-face interview weigert biedt vergelijking van de data van de face-to-face interviews met de postvragenlijsten een kijkje in de ‘blackbox’ van non-response. In onze studies waren geen significante non-response effecten. De tweede techniek is ontwikkeld om (mogelijke) problemen van representativiteit te ondervangen. Oftewel, iedere demonstrant heeft gelijke kans geïnterviewd te worden of om een vragenlijst aangeboden te krijgen. Beide demonstraties vonden plaats op een plein. Deze pleinen werden omsingeld door onderzoekers die vervolgens naar elk interview of vragenlijst 10 stappen naar het middelpunt van het plein liepen, waar vervolgens weer iemand werd geinterviewd of een vragenlijst werd uitgedeeld. Op deze wijze werden 2 maal SAMENVATTING 163 125 face-to-face interviews afgenomen en 2 maal 500 vragenlijsten uitgedeeld. Van deze 1000 post-vragenlijsten werden er 442 teruggestuurd (44 %). Dit is een goede response voor vragenlijsten zonder herinnering. In Hoofdstuk 4 werd het participatiemodel en de invloed van mobiliserende context getoetst (zie Figuur 7.1). Wij veronderstelden dat de demonstratie van de vakbonden meer power georiënteerd was en die van Keer het Tij meer value-georiënteerd. Een vergelijking van de twee demonstraties laat zien dat de twee directe routes naar motivatie gelijk zijn: zowel identiteit- als ideologie motieven hebben een direct effect op de motivatie. Echter, bij het value-georiënteerde protest van Keer het Tij was de invloed van ideologie motieven op motivatie om te participeren twee maal zo sterk las bij het power-georiënteerde protest. Figuur 7.1. Participatie routes naar politiek protest. identiteit instrumentaliteit woede motivatie ideologie Opvallend was dat instrumentaliteit geen direct effect op motivatie had als groeps-gebaseerde woede werd toegevoegd. Dit resultaat werpt licht op een belangrijke paradox die decenia lang de dicussie rondom het waarom van protest participatie heeft beheerst: waarom zou iemand zich inzetten voor het groepsbelang (lees waarom zou iemand kosten willen maken om de groep als geheel beter af te laten zijn)? Volgens Olson (1965) zijn ze in eerste instantie geneigd om de freerider optie te hanteren, immers een rationeel denkend groepslid laat uit kosten/baten overwegingen participatie over aan een ander. De omzetting van instrumentele motieven in groeps-gebaseerde woede zou kunnen verklaren waarom participanten met een instrumenteel motief tóch mee doen. Immers, het reguleren van emoties die een sociale of politieke misstand oproepen kan je alleen zelf doen doen, op het reguleren van emoties kan geen free ride worden genomen. SAMENVATTING 164 Hoewel de directe paden naar protest participatie hetzelfde zijn, verschillen de indirecte routes. Het belangrijkste verschil zat in de indirecte identiteitsroute, bij het powergeoriënteerde protest werden naarmate men zich sterker identificeerde met de bond de instrumentele motieven sterker wat zich vertaalde in groeps-gebaseerde woede terwijl sterkere identificatie met Keer het Tij tot sterkere ideologische motieven leiden wat zich vervolgens weer vertaalde in groeps-gebaseerde woede. In Hoofdstuk 4 werd ook gekeken of de motivationele constellatie van zwakke en sterke identificeerders verschilden. De Structural Equation Modellen (SEM) laten zien dat sterke of zwakke banden met een sociale beweging invloed hebben op de motivationele constellatie. Instrumentele, ideologische, en groeps-gebaseerde woede motieven van zwakke identificeerders worden niet beïnvloed door de organisatie. Het motivationele patroon van zwakke identificeerders lijkt gerelateerd aan de persoonlijke identiteit, zij ervaren grieven als een bedreiging van hun persoonlijke identiteit en emoties zijn niet groeps- maar individueel gebaseerd. Dit lijkt er op te wijzen dat de motivatie om te participeren in collectieve actie voortkomt uit de behoefte om persoonlijke belangen te beschermen (Wright et al., 1990) en het gevolg is van meer individualistische en opportunistische strategieën (Ellemers et al., 1999). De emoties en motieven van sterke identificeerders daarentegen, worden wel beïnvloedt door desterkebanden met de organisatie, dit suggereert dat een bedreiging van collectieve groepsbelangen resulteert in gedeelde grieven en emoties. Sterke identificeerders internaliseren groepsbelangen als eigenbelangen, dus motivatie om te participeren wordt versterkt door de motivatie om de groep te willen steunen en komt voort uit solidariteitsstrategieën (Ellemers et al., 1999). Ons onderzoek laat zien dat identificatie naast een direct effect ook een indirect effect heeft op de motivatie om te participeren in politiek protest. Dit maakt eens te meer de centrale rol van identificatie in protest gedrag duidelijk. In Hoofdstuk 5 werd tenslotte regulatieve focus toegevoegd aan ons participatie model. De eerste vraag die we in de hoofdstuk trachtten te beantwoorden was of een collective action frame in promotie termen overtuigender is voor promotors dan voor preventors en vice versa. Hiertoe werden eerst de demonstratie-oproepen van zowel de vakbonden als van Keer het Tij door 8 mensen gescoord in termen van promotie en preventie karakteristieken. Hieruit kwam naar voren dat de oproep van de vakbonden als meer preventie werd gezien en de oproep van Keer het Tij als meer promotie. Uitgaande van regulatieve fit veronderstelden en vonden wij dat er meer preventors met typisch preventie motieven (instrumentaliteit en identiteit) naar de vakbondsdemonstratie zouden komen en meer promotors met typisch promotie motieven (ideologie) naar de Keer het Tij demonstratie. SAMENVATTING 165 De laatste vraag die we wilden beantwoorden in dit proefschrift was of regulatieve focus fungeert als ‘besturingsmechanisme’, met andere woorden kan regulatieve focus voorspellen wie welke route neemt. Hiertoe werd promotie- en preventie-focus als onderliggend construct aan het participatie model toegevoegd. Dit laat zien dat preventors eerder geneigd zijn de instrumentele en identiteitsroute te nemen terwijl promotors eerder geneigd zijn de ideologie route te nemen. Dus regulatieve focus lijkt inderdaad te functioneren als besturingsmechanisme. Dit betekent dat regulatieve focus (tot op zekere hoogte) kan verklaren wie mee doet aan collectieve actie en waarom. De resultaten van de twee protest studies laten de meerwaarde zien van ons participatie model. De resultaten van de dual-paths modellen van zowel Simon en collega’s (instrumentele en identiteitsroute) alswel van Zomeren en collega’s (instrumentele en groeps-gebaseerde woede) werden gerepliceerd. Echter, ons integratief model verklaarde beduidend meer variantie dan enig ander voorgesteld model. Bovendien bleek het zinvol om het model uit te breiden met een ideologie route. Ook blijkt dat een participatiemodel waar zowel directe alswel indirecte effecten van identificatie zijn geïncorporeerd beter bij de data te passen en meer variantie te verklaren dan een model waar alleen de directe effecten geïncorporeerd zijn. Daarnaast laten de studies zien dat het zinvol is om regulatieve focus theorie toe te passen bij het verklaren van protest gedrag. Regulatieve fit laat zien waarom een bepaald collective action frame overtuigender is voor de één dan de ander terwijl regulatieve focus als besturingsmechanisme laat zien waarom mensen geneigd zijn een bepaalde route naar protest te nemen. Als reactie op de meer irrationele en emotionele verklaringen van protest gedrag uit de eerste helft van de vorige eeuw, verschoven academische analyses van protest gedrag naar meer rationele, structurele en organisationele verklaringen. Dit proefschrift laat zien dat zij hier in te ver zijn doorgeslagen. Uit onze resultaten komt naar voren dat afhankelijk van een interactie tussen individuele en organisationele kenmerken, demonstranten participeren op basis van zowel instrumentele, solidaire alswel expressieve overwegingen. Vanuit een rationeel mensbeeld geredeneerd zou ieder groepslid kiezen voor de freeriders optie. Echter, de werkelijkheid laat zien dat vele mensen keer op keer bereid zijnsoms zelfs tot de dood er op volgtom te participeren in politiek protest. Dit lijkt te suggereren dat toekomstig onderzoek naar protest gedrag vooruitgang kan boeken door een integratie van zowel de irrationele en emotionele verklaringen alswel de meer rationele, structurele en organisationele verklaringen. 166 References Aaker, J. L., & Lee, A. Y. (2001). "I" Seek Pleasures and "We" Avoid Pains: The Role of SelfRegulatory Goals in Information Processing and Persuasion. Journal of Consumer Research, 28, 33-49. Aelst, P. van, & Walgrave, S. (2001). Who is that (wo)man in the street? From the normalisation of protest to the normalisation of the protester. European Journal of Political Research, 39, 461-486. Aelst, P. van, Walgrave, S., & Decoster, K., (2000). Politiek wantrouwen en protest. In M. Hooghe (ed.). Sociaal kapitaal en democratie (p. 225- 253). Leuven: Acco. Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage. Allen, V. L., & Wilder, D. A. (1975). Categorization, belief similarity, and intergroup discrimination. Journal of Personality and Social Psychology, 32, 971-977. Arbuckle, J. L. (1997). AMOS user's guide. Version 3.6. Chicago: Smallwaters. Aspinwall, L. G. (2001). Dealing with adversity: Self-regulation, coping, adaptation, and health. In A. Tesser & N. Schwarz (Eds.), Blackwell Handbook of Social Psychology: Intraindividual Processes (p. 591-614). Malden, MA: Blackwell Publishers. Ayduk, O., May, D., Downey, G., & Higgins, E. T. (2003). Tactical Differences in Coping With Rejection Sensitivity: The Role of Prevention Pride. Personality and Social Psychology Bulletin, 29(4), 435-448. Babbie, E.R. (1990). Survey research methods.(2nd ed.). Belmont, CA: Wadsworth Publishing Company. Back, K. W. (1951). Influence through social communication. Journal of Abnormal and Social Psychology, 46, 9-23. Baddeley, A. D. (1979). Working memory and reading. In P. A. Kolers, M. E. Wrolstadt, & H. Bouma (Eds.), Processing of visible language (Vol. 1). New York: Plenum. Bandura, A. (1997). Self-efficacy: The exercise of control. New York: Freeman. Barnes, S. H., & Kaase, M. (Eds.) (1979). Political Action: Mass Participation in Five Western Democracies. Beverly Hills, CA: Sage. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173-1182. Berg, M. B., Janoff-Bulman, R., & Cotter, J. (2001). Perceiving Value in Obligations and Goals: Wanting to Do What Should Be Done. Personality and Social Psychology Bulletin, 27(8), 982-995. Bizman, A., Yinon, Y., & Krotman, S. (2001). Group-Based Emotional Distress: An Extension of Self-Discrepancy Theory. Personality and Social Psychology Bulletin, 27(10), 12911300. Blackwood, L., Lafferty, G., Duck, J., & Terry, D. (2003). Putting the group back into the unions: a social psychological contribution to understanding union support. The journal of industrial relations, 45(4), 485-504. Boekkooi, M. (2005). Stop de oorlog! De organisatoren en participanten van de antioorlogsdemonstratie van 15 februari 2003 (Stop the war! Organizers and participants of the anti-war demonstration of 15 February, 2003). Unpublished Master thesis, Vrije Universiteit, Amsterdam. Boninger, D. S., Krosnick, J. A., & Berent, M. K. (1995). Origins of Attitude Importance: SelfInterest, Social Identification, and Value Relevance. Journal of Personality and Social Psychology, 68(1), 61-80. Brewer, M. B., & Gardner, W. (1996). Who is this we? Levels of collective identity and self representations. Journal of Personality and Social Psychology, 71(1), 83-93. REFERENCES 167 Brewer, M. B., & Silver, M. D. (2000). Group Distinctiveness, Social identification, and Collective Action. In Stryker, S. Owens, T.J., & White, R.W. (Eds), Self, identity, and Social Movements (p. 153-171). Minneapolis: Minnesota Press. Britt, L. & Heise, D. (2000). From Shame to Pride in Identity Politics. In: Stryker, S. Owens, T.J., & White, R.W. (Eds), Self, Identity and Social Movements (p. 252-262). Minneapolis: University of Minnesota Press. Brewer, M. B., & Silver, M. D. (2000). Group Distinctiveness, Social identification, and Collective Action. In Stryker, S. Owens, T.J., & White, R.W. (Eds), Self, identity, and Social Movements (p. 153-171). Minneapolis: Minnesota Press. Brock, T.C. (1965). Communicator-recipient similarity and decision change. Journal of Personality and Social Psychology, 1, 650-654. Brockner, J., & Higgins, E. T. (2001). Regulatory Focus Theory: Implications for the Study of Emotions at Work. Organizational Behavior and Human Decision Processes, 86(1), 3566. Brown, R. (1980). Group processes: Dynamics within and between groups. Oxford:Basil Blackwell Byrne, B. M. (2004). Testing for multigroup invariance using AMOS graphics: a road less traveled. Structural Equation Modeling, 11(2), 272-300. Cacioppo, J. T., Petty, R. E., & Kao, C. E. (1984). The efficient assessment of need for cognition. Journal of Personality Assessment, 48, 306-307. Camacho, C. J., Higgins, E. T., & Luger, L. (2003). Moral Value Transfer From Regulatory Fit: What Feels Right Is Right and What Feels Wrong Is Wrong. Journal of Personality and Social Psychology, 84(3), 498-510. Carver, C. S., & White, T. L. (1994). Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: The BIS/BAS Scales. Journal of Personality and Social Psychology, 67, 319-333. Cesario, J., Grant, H., & Higgins, E. T. (2004). Regulatory Fit and Persuasion: Transfer From "Feeling Right". Journal of Personality and Social Psychology, 86(3), 388-404. Cialdini, R. B., Trost, M. R., & Newsom, J. T. (1995). Preference for Consistency: The Development of a Valid Measure and the Discovery of Surprising Behavioral Implications. Journal of Personality and Social Psychology, 69(2), 318-328. Deci, E. L., & Ryan, R. M. (1991). A motivational approach to the self: Integration in personality. In R. A. Dienstbier (Ed.), Perspectives on motivation (p. 237-288). Nebraska: University of Nebraska Press. Demeester, E. (2004). Reawakening of the Dutch working class. Retrieved 19-4-2005, from http://www.marxist.com/Europe/netherlands_oct_demo.htm Deutsch, M. & Gerard, H. B. (1955). A study of normative and information social influences upon individual judgement. Journal of Abnormal and Social Psychology, 51, 629-636. Devos, T., Silver, L. A., & Mackie, D. M. (2002). Experiencing intergroup emotions. In D. M. Mackie & E. R. Smith (Eds.), From Prejudice to Intergroup Emotions: Differentiated Reactions to Social Groups (p. 111-134). Philadelphia, PA: Psychology Press. Drury, J. & Reicher, S. (1999). The intergroup dynamics of collective empowerment: Substantiating the social identity model. Group Processes and Intergroup Relations, 2, 381-402. Drury, J. & Reicher, S. (2005). Explaining enduring empowerment: A comparative study of collective action and psychological outcomes. European Journal of Social Psychology, 35, 35-58. Ekman, P. (1993) Facial expression and emotion. American Psychologist, 48, 384-392. Ellemers, N., Spears, R., & Doosje, B. (1999). Introduction. In N. Ellemers, R. Spears, & B. Doosje (Eds.), Social identity: Context, Commitment, Content. (p. 1-5). Oxford: Basil Blackwell. REFERENCES 168 Elliot, A. J., & Thrash, T. M. (2002). Approach-avoidance motivation in personality: Approach-avoidance temperaments and goals. Journal of Personality and Social Psychology, 82, 804-818. Elliot, A. J., & Thrash, T. M. (2004). The Intergenerational Transmission of Fear of Failure. Personality and Social Psychological Bulletin, 30(8), 957-971. Erkut, S., Alarcón, O., García Coll, C., Tropp, L. R., & Vázquez García, H. A. (1999). The dualfocus approach to creating bilingual measures. Journal of Cross Cultural Psychology, 30(2), p. 206-218. Fazio, R. H. (1995). Attitudes as object-evaluation associations: determinants, consequences, and correlates of attitude accessibility. In R. E. Petty, & J. A. Krosnick (Eds.), Attitude strength: Antecedents and consequences (p. 247–282). Hillsdale, NJ: Erlbaum. Feather, N.T., & Newton, J.W. (1982). Values, expectations, and the prediction of social action: An expectancy-valence analysis. Motivation and Emotion, 6, 217-244. Feldman, S. (2003). Values, Ideology, and the Structure of Political Attitudes. In D. Sears, L. Huddy, & R. Jervis (Eds.), Oxford Handbook of Political Psychology (p. 477-508). New York: Oxford University Press. French, J. R. P., & Raven, B. H. (1959). The bases of social power. In D. Cartwright (Ed.), Studies in social power (p. 150-167). Ann Arbor: University of Michigan. Gamson, W.A. (1992). The Social Psychology of Collective Action. In A.D. Morris and C.M. Mueller (Eds.). Frontiers in Social Movement Theory (p. 53-76). New Haven, CT: Yale University Press. Gray, J. A. (1994). Three fundamental emotion systems. In P. Ekman & R. J. Davidson (Eds.), The nature of emotion (p. 243–247). New York: Oxford University Press. Guadagno, R. E., Asher, T., Demaine, L. J., & Cialdini, R. B. (2001). When Saying Yes Leads to Saying No: Preference for Consistency and the Reverse Foot-in-the-Door Effect. Personality and Social Psychological Bulletin, 27, 859-867. Guay, F., Vallerand, R. J., & Blanchard, C. (2000). On the assessment of situational intrinsic and extrinsic motivation: The Situational Motivation Scale (SIMS). Motivation and Emotion, 24(3), 175-213. Goodwin, J. & Jasper, J. M. (1999). Caught in a Winding, Snarling Vine: The Structural Bias of Political Process Theory. Sociological Forum,14, 27-54. Goodwin, J., Jasper, J. M., & Polleta, F. (2001). Passionate Politics. Emotions and Social Movements. Chicago: The University of Chicago Press. Harmon-Jones, E., Sigelman, J. D., Bohlig, A., & Harmon-Jones, C. (2003). Anger, coping, and frontal cortical activity: The effect of coping potential on anger-induced left frontal activity. Cognition and Emotion, 17, 1-24. Higgins, E. T. (1987). Self-Discrepancy: A Theory Relating Self and Affect. Psychological Review, 94(3), 319-340. Higgins, E. T. (1996). The "Self Digest": Self-Knowledge Serving Self-Regulatory Functions. Journal of Personality and Social Psychology, 71(6), 1062-1083. Higgins, E. T. (1997). Beyond Pleasure and Pain. American Psychologist, 52(12), 1280-1300. Higgins, E. T. (1998). Promotion and prevention: Regulatory focus as a motivational principle. In: M. P. Zanna (Ed.), Advances in experimental social psychology (p. 1–46). New York: Academic Press. Higgins, E. T. (2000). Making a Good Decision: Values From Fit. American Psychologist, 55(11), 1217-1230. Higgins, E. T. (2002). How Self-Regulation Creates Distinct Values: The Case of Promotion and prevention decision Making. Journal of Consumer Psychology, 12, 177-191. Higgins, E.T., Bond, R.N., Klein, R., & Strauman, T. (1986). Self-discrepancies and emotional vulnerability: how magnitude, accessibility, and type of discrepancy influence affect. Journal of Personality and Social Psychology, 51, 5-51. REFERENCES 169 Higgins, E. T., Friedman, R. S., Harlow, R. E., Idson, L. C., Ayduk, O. N., & Taylor, A. (2001). Achievement Orientations from Subjective Histories of Success: Promotion Pride versus Prevention Pride. European Journal of Social Psychology, 31, 3-23. Higgins, E. T., Idson, L. C., Freitas, A. L., Spiegel, S., & Molden, D. C. (2003). Transfer of Value From Fit. Journal of Personality and Social Psychology, 84(6), 1140-1153. Higgins, E. T., Klein, R., & Strauman, T. (1985). Self concept discrepancy theory: A psychological model for distinguishing among different aspects of depression and anxiety. Social Cognition, 3, 51-76. Higgins, E. T., Shah, J., & Friedman, R. (1997). Emotional responses to goal attainment: strength of regulatory focus as moderator. Journal of Personality and Social Psychology, 72(3), 515-525. Higgins, E. T., & Silberman, I. (1998). Development of regulatory focus: Promotion and prevention as ways of living. In J. Heckhausen & C. S. Dweck (Eds.), Motivation and self-regulation across the life span (p. 78–113). New York: Cambridge Univ. Press. Hook, E. V., & Higgins, E. T. (1988). Self-related Problems Beyond the Self-Concept: Motivational Consequences of Discrepant Self-guides. Journal of Personality and Social Psychology, 55(4), 625-633. Huddy, L. (2001). From Social to Political Identity: A Critical Examination of Social Identity Theory. Political Psychology, 22, 127-156. Huddy, L. (2003). Group Identity and Political Cohesion. In D. O. Sears, L. Huddy & R. Jervis (Eds.), Oxford handbook of Political psychology (p. 511-558). Oxford: University Press. Hu, L.-t., & Bentler, P. M. (1999). Cutoff Criteria for Fit Indexes in Covariance Structure Analysis: Conventional Criteria Versus New Alternatives. Structural Equation Modeling, 6(1), 1-55. Inglehart, R. (1987) Value Change in Industrial Societies. American Political Science Review, 81, 1289-1303. Inglehart, R. & Catterber, G. (2002). Trends in Political Action. The Developmental Trend and the Post honeymoon Decline. International Journal of Comparative Sociology 43(3). Jasper, J. M. (1997). The Art of Moral Protest. Culture, Biography, and Creativity in Social Movements. Chicago: The University of Chicago Press. Jasper, J. M. (1998). The Emotions of Protest: Affective and reactive Emotions In and Around Socoal Movements. Sociological Forum, 13(3), 397-424. Kelly, C., & Breinlinger, S. (1996). The social psychology of collective action: Identity, injustice and gender. London: Taylor & Francis. Kelly, J & Kelly, C (1994). Who gets involved in collective action? Social psychological determinants of individual participation in trade unions. Human Relations 47(1), 63-88. Kim, H. (2002). Shame, Anger, and Love in Collective Action: Emotional consequences of Suicide Protest in South Korea, 1991. Mobilization: An International Journal, 7(2), 159176. Klandermans, B. (1984). Mobilization and participation: Social psychological expansions of resource mobilization theory. American Sociological Review, 49, 583-600. Klandermans, B. (1988). The formation and mobilization of consensus. International Social Movement Research, 1, 173-196. Klandermans, B. (1993). A theoretical framework for comparisons of social movement participation. Sociological forum. 8, 383-402. Klandermans, B. (1997). The social psychology of protest. Oxford: Basil Blackwell. Klandermans, B. (2000). Identity and Protest: How group identification helps to overcome collective action dilemmas. In Vugt, van, M., Snyder, M., Tyler, T. R., & Biel, A. (Eds), Cooperation in modern society: Promoting the welfare of communities, states, and organizations (p. 162-183). Edited by London: Routledge. REFERENCES 170 Klandermans, B. (2003). Collective Political Action. In D. O. Sears, L. Huddy, & R. Jervis (Eds.), Oxford handbook of Political Psychology (p. 670-709). Oxford: University Press. Klandermans, B. (2004). The demand and Supply of Participation: Social-Psychological Correlates of Participation in Social Movements. In D. A. Snow, S. A. Soule, & H. Kriesi (Eds.), The Blackwell companion to social movements (p. 360-379). Oxford: Blackwell Publishing. Klandermans, B., & Oegema, D. (1987). Potentials, networks, motivations, and barriers: Steps towards participation in social movements. American Sociological Review, 52, 519-531. Klandermans, B., & Smith, J. (2002). Survey Research: A Case for Comparative Design. In B. Klandermans & S. Staggenborg (Eds.), Methods of Social Movement Research (p. 3-31). Minneapolis, MN: the University of Minnesota Press. Klandermans, B. & J. Stekelenburg, van (forthcoming). Instrumentality, Identity, and Groupbased Anger: Social Psychological Approaches to Protest Participation. Landis, J. R., & Koch, G. G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33, 159–174. Lang, P. J. (1995). The emotion probe: Studies of motivation and attention. American psychologist, 50, 372-385. Lazarus, R. S. (1991). Emotion and adaptation. New York, NY, Oxford University Press. Lazarus, R. S. (2001). Relational meaning and discrete emotions. In K. R. Scherer, A. Schorr, & T. Johnstone (Eds.), Appraisal processes in emotion (p. 37–67). Oxford, England: Oxford University Press. Lazarus, R.S. &: Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer Publishing Company. Levine, J. M., Higgins, E. T., & Choi, H-S. (2000). Development of strategic norms in groups. Organizational Behavior and Human Decision Processes, 82, 88-101. Lee, A. Y., Aaker, J. L., & Gardner, W. L. (2000). The Pleasures and Pains of Distinct SelfConstruals: The Role of Interdependence in Regulatory Focus. Journal of Personality and Social Psychology, 78(6), 1122-1134. Lockwood, P., Jordan, C. H., & Kunda, Z. (2002). Motivation by Positive or Negative Role Models: Regulatory Focus Determines Who Will Best Inspire Us. Journal of Personality and Social Psychology, 83(4), 854-864. Lockwood, P., Marshall, T. C., & Sadler, P. (2005). Promoting Success or Preventing Failure: Cultural Differences in Motivation by Positive and Negative Role Models. Personality and Social Psychology Bulletin, 31(3), 379-392. Lockwood, P., Sadler, P., Fyman, K., & Tuck, S. (2004). To Do or Not to Do: Using Positive and Negative Role Models to Harness Motivation. Social Cognition, 4, 422-450. Mackie, D. M. (1986). Social identification effects in group polarization. Journal of Personality and Social Psychology, 50, 720-728. Mackie, D.M. & Cooper, J. (1984). Group polarization: The effects of group membership. Journal of Personality and Social Psychology, 46, 575-585. Mackie, D. M., Devos, T., & Smith, E. R. (2000). Intergroup emotions: Explaining offensive action tendencies in an intergroup context. Journal of Personality and Social Psychology, 79, 602-616. Mackie, D. M., Gastardo-Conaco, M. C., & Skelly, J. J. (1992). Knowledge of the advocated position and the processing of in-group and out-group persuasive messages. Personality and Social Psychology Bulletin, 18, 145-151. Mackie, D. M., Worth, L. T., & Asuncion, A. G. (1990). Processing of persuasive in-group messages. Journal of Personality and Social Psychology, 58, 812-822. McAdam, D. & Zald, M (Eds.). (1996). Comparative perspectives on Social Movements. Chicago: Cambridge University Press. REFERENCES 171 McCarthy, J. D., McPhail, C., & Smith, J. (1996). Images of Protest. American Sociological Review, 61, 478-499. McCoy, S. K. & Major, B. (2003). Group identification moderates emotional responses to perceived prejudice. Personality and Social Psychology Bulletin, 29(8), 1005-1017. McKeachie, W. & Doyle, C., (1970). Psychology. Reading, MA: Addison-Wesley. Mieder, W. (1995). "The grass is always greener on the other side of the fence": an American proverb of discontent. DeProverbio, 1(1). Miller, D. T. (2000). On the experience of injustice. Annual Review of Psychology, 52, 527–553. Mischel, W., Cantor, N., & Feldman, S. (1996). Principles of self-regulation: The nature of willpower and self-control. In E. T. Higgins & A. Kruglanski (Eds.), Social psychology: Handbook of basic principles (p. 329-360). New York: Guilford. Molden, D. C., & Higgins, E. T. (2004). Categorization under uncertainty: resolving vagueness and ambiquity with eager versus vigilant strategies. Social Cognition, 22(2), 248-277. Mummendey, A., Kessler, T., Klink, A., & Mielke, R. (1999). Strategies to cope with negative social identity: Predictions by social identity theory and relative deprivation theory. Journal of Personality and Social Psychology, 76, 229-245. Neuberg, S. L. & Newsom, J. T. (1993). Personal need for structure - individual-differences in the desire for simple structure. Journal of Personality and Social Psychology, 65(1), 113131. Norris, P. (2002). Democratic Phoenix: Political Activism worldwide. New York: Cambridge University Press. Norris, P. (2004). Political activism: New challenges, new opportunities. C. Boix & S. Stokes (Eds.). Oxford Handbook of Comparative Politics, Oxford: University Press. Olson, M. (1965). The logic of collective action: public goods and the theory of groups. Cambridge, MA: Harvard University Press. Norris, P., Walgrave, S., & van Aelst, P. (2005). Who Demonstrates? Antistate Rebels, Conventional Participants, or Everyone? Comparative Politics, 37(2), 189-205. Omoto, A. M. & Snyder, M. (1993). AIDS volunteers and their motivations: Theoretical issues and practical concerns. Nonprofit Management and Leadership, 4, 157-176. Perkins, F.M.; Werner, M.U.; Persson, F.; Holte, K.; Jensen, T.S. & Kehlet, H. (2004). Development and validation of a brief, descriptive Danish pain questionnaire (BDDPQ). Acta Anaesthesiologica Scandinavica, 48(4), 486-490. Pham, M. T. & Avnet, T. (2004). Ideals and oughts and the reliance on affect versus substance in persuasion. Journal of Consumer Research, 30, 503–518. Pham, M. T. & Higgins, E. T. (2005). Promotion and prevention in consumer decision making: State of the art and theoretical propositions. In S. Ratneshwar & D. G. Mick (Eds.), Inside consumption: Perspectives on consumer motives, goals, and desires (p. 8–43). London: Routledge. Pieters, R. G. M., Verplanken, B., & Modde, J. M. (1987). 'Neiging tot nadenken': samenhang met beredeneerd gedrag. Nederlands Tijdschrift voor de Psychologie en haar grensgebieden, 42(2), 62-66. Polletta, F. (1999). Snarls, Quacks, and Quarrels: Culture and Structure in Political Process Theory. Sociological Forum, 14, 67-74. Putnam, R. (2000). Bowling Alone: The Crumbling and Revival of American Community, New York: Simon & Schuster. Reichenbach, R. (2000). The scandalised self. Paper presented at the Association of Moral Education (AME), Glasgow, Scotland. 7/11-7-2000. Reicher, S. (1984) The St. Paul's riot: An explanation of the limits of crowd action in terms of a social identity model. European Journal of Social Psychology, 14, 1-21. Reicher, S. (2004). The Context of Social Identity: Domination, Resistance, and Change. Political Psychology, 25(6), 921-945. REFERENCES 172 Reicher, S. (2005). Becoming the subjects of history: new directions for the psychology of collective action. Talk delivered at Workshop on Collective Identities, Mobilization and Political Participation Kiel (Germany), September 1st – 4th 2005. Rij, C., van (1995). Naar de bond: Vakbondsloopbanen en beroepsloopbanen. In Klandermans, B. & Visser, J. (Eds.), De vakbeweging na de welvaartsstaat (p. 67-86). Assen: Van Gorcum. Rokeach, M. (1973). The nature of human values. New York: Free Press. Roggeband, C.M. (2002) Over de grenzen van de politiek. Een vergelijkend onderzoek naar de opkomst en ontwikkeling van de vrouwenbeweging tegen seksueel geweld in Nederland en Spanje. Assen: Van Gorcum. Rokeach, M. (1973). The nature of human values. New York: Free Press. Roller, E. & Wessels, B. (1996). Contexts of political protest in Western democracies: political organization and modernity. In: Weil, F.D. (Ed), Extremism, Protest, Social Movements and democracy. Greenwhich: JAI Press. Ronen, S. (1994). An underlying structure of motivational need taxonomies: A cross cultural confirmation. In M. D. Dunnette & L. M. Hough (Eds.), Handbook of Industrial and organizational psychology (Vol. 4). Palo Alto, CA: Consulting Psychology. Rucht, D., Koopmans, R., & Neidhardt, F. (1998). Introduction: Protest as a Subject of Empirical Research. In D. Rucht, R. Koopmans, & F. Neidhardt (Eds), Acts of Dissent: NewDevelopments in the Study of Protest (p. 7-32). Berlin: Edition Sigmas Rainer Bohn Verlag. Rubin, J. Z., Pruitt, D. G., & Kim, S. H. (1994). Social conflict: Escalation, stalemate, and settlement (2nd ed.). New York: McGraw-Hill. Sassenberg, K., Kessler, T., & Mummendey, A. (2003). Less negative=more positive? Social discrimination as avoidance or approach. Journal of Experimental Social Psychology, 39(1), 48-58. Scherer, K. R. & Zentner, M. R. (2001). Emotional effects of music production rules. In: Juslin, P.N. & Sloboda, J.A. (Eds.). Music and emotion: theory and research (p. 361-392). Oxford: University Press. Schrager, L.S. (1985). Private Attitudes and Collective Action. Comment on Klandermans, ASR, October. 1984. American Sociological Review, 50, 858-859. Schwartz S. H. (1992). Universals in the content and structure of values: theoretical advances and empirical tests in 20 countries. In M. P. Zonna (Ed.), Advances in Experimental Social Psychology (p. 1-65). San Diego: Academic Press. Scott, A. (1990). Ideology and the New Social Movements. London: Routledge. SCP. (1999). the Amenities and Services Utilisation Survey (AVO) 1999. Retrieved 4 September, 2004 SCP. (2004). Summary of the social state of The Netherlands. The Hague: Social and Cultural Planning Office (SCP working paper 107, English edition). Shah, J. Y., Brazy, P. C., & Higgins, E. T. (2004). Promoting Us or Preventing Them: Regulatory Focus and Manifestations of Intergroup Bias. Personality and Social Psychology Bulletin, 30 (4), 433-446. Shah, J. & Higgins, E. T. (1997). Expectancy × Value effects: regulatory focus as determinant of magnitude and direction. Journal of Personality and Social Psychology, 73(3), 447-458. Shah, J. Y. & Higgins, E. T. (2001). Regulatory concerns and appraisal efficiency: The general impact of promotion and prevention. Journal of Personality and Social Psychology, 80, 693-705. Simon, B. (1999). A place in the social world: Self and social categorization. In T. R. Tyler, R. M. Kramer, & O. P. Johns (Eds.), The Psychology of the Social Self. (p. 47-69). Mahwah, NJ: Erlbaum. Simon, B. & Klandermans, B. (2001). Politicized collective identity. American Psychologist, 56(4), 319-331. REFERENCES 173 Simon, B., Loewy, M., Stürmer, S., Weber, U., Freytag, P., Habig, C., et al. (1998). Collective Identification and Social Movement Participation. Journal of Personality and Social Psychology, 74(3), 646-658. Smith, E. R. (1993). Social identity and social emotions: Toward new conceptualizations of prejudice. In D. M. Mackie & D. L. Hamilton (Eds.), Affect, cognition, and stereotyping: Interactive processes in group perception (p. 297–315). San Diego, CA: Academic Press. Smith, E. R. (1999). Affective and cognitive implications of a group becoming part of the self: New models of prejudice and of the self-concept. In M. A. Hogg & D. Abrams (Eds.), Social identity and social cognition (p. 183-196). Oxford, England: Basil Blackwell. Snow, D.A. & Benford, R.D. (1988). Ideology, Frame Resonance and Participant Mobilization. International Social Movement Research, 1, 197-219. Snow, D., Rochford, B. B., Worden, S. K., & Benford, R. D.. (1986). Frame alignment processes, micromobilization, and movement participation. American Sociological Review, 51, 46481. Snow, D. A., Zurcher, L., A., Jr., & Ekland-Olson, S. (1980). Social networks and social movements: A microstructural approach to differential recruitment. American sociological review, 45(5), 787-801. Stekelenburg, J. v. & Klandermans, B. (2003). Regulatory focus meten met behulp van spreekwoorden (Using proverbs to measure regulatory focus). In D. Wigboldus, M. Dechesne, E. Gordijn & E. Kluwer (Eds.), Jaarboek Sociale Psychologie 2003 (JSP) (p. 345-357). Groningen: ASPO pers. Stekelenburg, J. v. & Klandermans, B. (2004). Hebben promotie- en preventiegerichte demonstranten andere participatiedoelen en emoties? (Employ promotion- and prevention-focused protesters different participation goals and do they experience different emotions?) In E. H. Gordijn, R. Holland, A. Meijnders, & J. Ouwerkerk (Eds.), Jaarboek Sociale Psychologie 2004 (JSP) (p. 371-380). Groningen: ASPO pers. Strauman, T. J. (1996). Stability within the self: A longitudinal study of the structural implications of self-Discrepancy theory. Journal of Personality and Social Psychology, 71(6), 1142-1153. Stroessner, S.J. (2005). Threat, Vigilance, & Stereotyping Research program. Retrieved 15-102005 from http://www.barnard.edu/psych/faculty/Stroessner/Research.htm. Stryker, S., Owens, T. J., & White, R. W. (Eds.), (2000). Self, Identity, and Social Movements. Minneapolis: Minnesota Press. Stürmer, S. & Kampmeier, C. (2003). Active citizenship: The role of community identification in community volunteerism and local participation. Psychologica Belgica, 43, 103 – 122. Stürmer, S. & Simon, B. (2004). Collective action: Towards a dual-pathway model. In W. Stroebe, & M. Hewstone (Eds.), European review of social psychology (p. 59-99). Hove, UK: Psychology Press. Stürmer, S., Simon, B., Loewy, M., & Jörger, H. (2003). The dual-pathway model of social movement participation: The case of the fat acceptance movement. Social Psychology Quarterly, 66, 71 – 82.! Sverke, M. (1996). The importance of ideology in trade union participation in Sweden: Asocialpsychological model. In Pasture, P., Verberckmoes, J., & De Witte, H. (Eds.), The lost perspective: Trade unions between ideology and social action in the new Europe (p. 353– 376). Aldershot: Avebury. Sverke, M., & Sjöberg, A. (1997). Ideological and instrumental union commitment. In M. Sverke (Ed.), The future of trade unionism: International perspectives on emerging union structures (p. 277–293). Aldershot: Ashgate. Tangney, J. P., Niedenthal, P. M., Covert, M. V., & Barlow, D. H. (1998). Are shame and guilt related to distinct self-discrepancies? A test of Higgins's (1987) hypotheses. Journal of Personality and Social Psychology, 75(1), 256-268. REFERENCES 174 Terry, D. J. & Hogg, M. A. (1996). Group norms and the attitude-behavior relationship: A role for group identification. Personality and Social Psychology Bulletin, 22, 776–793. Tyler, T.R. & Blader, S.L. (2003). The Group Engagement Model: Procedural Justice, Social Identity, and Cooperative Behavior. Personality and Social Psychology Review, 7 (4), 349-361. Tajfel, H. & Turner, J. (1979). An integrated theory of intergroup conflict. In W. C. Austin & S. Worchel (Eds.), The Social Psychology of Intergroup Relations. Montery, CA: Brooks Cole. Turner, J. C. (1999). Some current issues in research on social identity and self-categorization theories. In N. Ellemers, R. Spears, & B. Doojse (Eds.), Social identity: Context, commitment, content (p. 6–34). Oxford: Blackwell. Turner, R. H. & Killian, L. M. (1987). Collective behavior (3rd ed.). Englewood Cliffs, NJ: Prentice-Hall. Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the social group: A self-categorization theory. Oxford: Blackwell. Van-Dijk, D. & Kluger, A. N. (2004). Feedback Sign Effect on Motivation: Is it Moderated by Regulatory Focus? Applied Psychology An International Review, 53(1), 113-153. Veen, G.,van der (1992). Principes in praktijk. Kampen: Kok. Veenstra, K. & Haslam, S. A. (2000). Willingness to participate in industrial protest: Exploring social identification in context. British Journal of Social Psychology, 39(2), 153-172. Walgrave, S. (2005). Understanding Political Protest. Protest surveying in a comparative perspective. Research proposal. University of Antwerp, Belgium. Walonick, D. S. (1997). Survival Statistics. Bloomington, MN: StatPac. Walsh, E. J. (1988). Democracy in the Shadows: Citizen Mobilization in the Wake of the Accident at the Three Mile Island, New York: Greenwood Press. Wederspahn, G. M. (2002). Proverbs: Windows Into Other Cultures. Retrieved 03-07-2003, from http://www.executiveplanet.com/business-culture/ Weerd, M., de & Klandermans, B. (1999). Group identification and political protest: farmers' protest in the Netherlands. European Journal of Social Psychology, 29, 1073-1095. Welzel, C., Inglehart, R., & Deutch, F. (2005). Social Capital, Associational Life, and Collective Action. What Makes Some Publics more Civic than others? Bremen University Workpapers (www.iu-bremen.de/schools/shss/research/cis/papers). Wilder, D. A. (1990). Some determinants of the persuasive power of in-group and out-groups: Organization of information and attribution of independence. Journal of Personality and Social Psychology, 59, 1202-1213. Wilder, D.A. & Shapiro, P.N. (1984). Roles of out-group cues in determining social identity. Journal of Personality and Social Psychology, 47(2), 342-348. Witte, H.,de. (1995), ‘Worden vakbondsleden (nog) bewogen door ideologie?’, Tijdschrift voor Arbeidsvraagstukken, 11 (3), 261-279. Wright, S. C., Taylor, D. M., & Moghaddam, F. M. (1990). Responding to membership in a disadvantaged group: From acceptance to collective protest. Journal of Personality and Social Psychology,58, 994-1003. Yzerbyt, V., Dumont, M., Wigboldus, D., & Gordijn, E. (2003). I feel for us: The impact of categorization and identification on emotions and action tendencies. British Journal of Social Psychology, 42(4), 533-549. Zomeren, M., van, Spears, R., Fischer, A. H., & Leach, C. W. (2004). Put Your Money Where Your Mouth Is! Explaining Collective Action Tendencies Through Group-Based Anger and Group Efficacy. Journal of Personality and Social Psychology, 87(5), 649-664. Zajonc, R.B. (1966). Social psychology: An experimental approach. Belmont, California: Brooks/Cole. 175 The “Kurt Lewin Institute Dissertation Series” started in 1997. Since 2004 the following dissertations have been published: 2004-1: 2004-2: 2004-3: 2004-4: 2004-5: 2004-6: 2004-7: 2004-8: 2004-9: 2004-10: 2004-11: 2004-12: 2004-13: 2004-14: 2004-15: 2005-1: 2005-2: 2005-3: 2005-4: 2005-5: 2005-6: 2005-7: 2005-8: 2005-9: 2005-10: 2005-11: 2005-12: 2005-13: 2005-14: 2005-15: Wolfgang Steinel: Misleading in social decision-making: A motivational approach Edwin van Hooft: Job seeking as planned behavior. In search of group differences Kirsten Ruys: The influence of self and similarity on immediate affective impressions of people Michel Handgraaf: Self-interest and other-oriented motives in social decision making: An ultimatum bargaining perspective Sjoerd Goslinga: Betrokkenheid, participatie en ledenverloop in vakbonden Nynke Frieswijk: Frail, but happy: The importance of self-management ability and social comparison among elderly persons Jaap Ham: Bridging attribution and spontaneous inferences: Spontaneous and intentional components of dispositional and situational inferences Mirjam Tazelaar: When things go wrong: The influence of noise on interactions Marjolein de Best-Waldhober: Coördinatie in interdependente situaties: Over de afweging van directe uitkomsten versus uitkomsten op lange termijn Hinke Groothof: When others are doing better or worse: Responses from the heart and the head Laetitia Mulder: Throwing light on the dark side of sanctions. Sanctioning systems in social dilemmas re-examined Annebel de Hoogh: Personality and charismatic leadership: A matter of context Brigitte ten Brink: Psychological contract: A useful concept? Gerben van Kleef: Emotion in social conflict. The interpersonal effects of emotions in negotiations Tom Frijns: Keeping secrets: Quantity, quality and consequences Elsbeth Steenland: Stereotype Change: Effects of disconfirming information Annet de Lange: What about causality? Examining longitudinal relations between work characteristics and mental health Reint Jan Renes: Sustained volunteerism: Justification, motivation and management Aloyse Augusta Dinsbach: Socialization in the workplace: A focus on migrant groups Susanne Peters: The social psychology of being better off than others Winny Bakker: Emigration and well-being: The role of personality and cultural identity in acculturation Ruud Zaalberg: The expression of emotion in social situations. The mediating roles of display rules and social motives Catharine Evers: Sex differences in anger expressions. The shaping role of social appraisals Ed Sleebos: The Consequences of Perceived Intra-group Respect: The Effects of Differential Intra-group Respect on Behavior and Cognition Jeroen Stouten: Virtue Summons the Fury: Coordination Rules and Fairness in Social Dilemmas Irene de Pater: Doing Things Right or Doing the Right Thing: A New Perspective on the Gender Gap in Career Success Rob Nelissen: Guided by Reason, Struck by Emotion: Integrating Motivational & Expectancy - Value Accounts of Behavior Emely de Vet: Testing the Transtheoretical Model: Validity and Applicability for Fruit Intake Floor Rink: Diversity and Small Group Decision Making: Towards a social identity framework for studying the effects of task-related differences on dyads and groups Sven Zebel: Negative Associations: The Role of Identification in Group-Based Guilt 176 2005-16: Eric Rietzschel: From quantity to quality: Cognitive, motivational and social aspects of creative idea generation and selection 2006-1: Maria Dijkstra: Workplace Conflict and Individual Well-Being 2006-2: Ruud Custers: On the underlying mechanisms of nonconscious goal pursuit 2006-3: Ellen Dreezens: The missing link: the relationship between values and attitudes 2006-4: Jacquelien van Stekelenburg: Promoting or preventing social change. Instrumentality, identity, ideology and groupbased anger as motives of protest participation

RELATED PAPERS

RELATED TOPICS

Log In

Promoting or Preventing Social Change

Promoting or Preventing Social Change

Related Papers

RELATED PAPERS

RELATED TOPICS