Book Chapter The biographical impact of participation in social movement activities: beyond highly committed New Left activism GIUGNI, Marco, GRASSO, Maria T. Abstract Studying the outcomes of social movements is important if we want to elucidate the role of collective action in society. While most works have addressed aggregate-level political outcomes such as changes in laws or new policies, a relatively small but substantial body of literature deals with the personal and biographical consequences of social movements at the micro-level, that is, effects on the life-course of individuals who have participated in movement activities, due at least in part to involvement in those activities (see Giugni 2004 for a review). In general, these studies converge in suggesting that activism has a strong effect both on the political and personal lives of the subjects. Most of the existing studies, however, share a number of features that limit the scope of their findings. First, they focus on a specific kind of movement participants, namely movement activists, most often New Left activists, who are strongly committed to the cause. Yet, as suggested by McAdam (1999a), not more than between 2% and 4% of the American population took part in New Left activism of the 1960s. As a result, we cannot [...] Reference GIUGNI, Marco, GRASSO, Maria T. The biographical impact of participation in social movement activities: beyond highly committed New Left activism. In: Lorenzo Bosi, Marco Giugni & Katrin Uba. The Consequences of Social Movements. Cambridge : Cambridge University Press, 2016. p. 88-105 Available at: http://archive-ouverte.unige.ch/unige:85872 Disclaimer: layout of this document may differ from the published version. C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 4 85 [85–105] 15.9.2015 1:59PM THE BIOGRAPHICAL IMPACT OF PARTICIPATION IN SOCIAL MOVEMENT ACTIVITIES Beyond highly committed New Left activism Marco Giugni and Maria T. Grasso Introduction Studying the outcomes of social movements is important if we want to elucidate the role of collective action in society. While most works have addressed aggregate-level political outcomes such as changes in laws or new policies, a relatively small but substantial body of literature deals with the personal and biographical consequences of social movements at the micro-level, that is, effects on the life-course of individuals who have participated in movement activities, due at least in part to involvement in those activities (see Giugni 2004 for a review). In general, these studies converge in suggesting that activism has a strong effect both on the political and personal lives of the subjects. Most of the existing studies, however, share a number of features that limit the scope of their findings. First, they focus on a specific kind of movement participants, namely movement activists, most often New Left activists, who are strongly committed to the cause. Yet, as suggested by McAdam (1999a), not more than between 2% and 4%S of the American population took part in New Left activism of the 1960s. As a result, we cannot directly generalize from these findings to the biographical impact of participation in We thank Lorenzo Bosi and Katrin Uba, as well as the participants of the MOVEOUT workshop held at Uppsala University on September 6–8, 2012, for the useful comments on previous drafts of this chapter. We also thank James Laurence and Alessandro Nai for the advice with the data analysis. C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 86 [85–105] 15.9.2015 1:59PM 86 / Marco Giugni and Maria T. Grasso social movements by less strongly committed people who, in addition, might belong to other ideological areas, not necessarily to the New Left. Second, they use only a limited number of subjects and often do not analyze non-activists. The possibility to generalizing the findings are therefore very limited due to the small samples used and the lack of a control group of non-activists. Third, they look at a specific geographical and area and historical period, namely the United States (or often even more circumscribed geographical areas) in the late 1960s and early 1970s.As such, we do not know from these studies how movement participation may affect the lives of people more generally. Thus, overall, in spite of the crucial insights that these works provided, they have little to say about the effects of more “routine” forms of participation. In this chapter we try to go beyond the traditional focus on highly committed New Left activism to investigate the impact of protest participation on political life-course patterns amongst the general population. Our main research question is the following: Does participation in social movement activities, such as participation in protest activities, have an enduring impact on the subsequent political life of individuals? This question is linked to two more specific research questions: Does run-of-the-mill activism have an effect on the political attitudes and behaviors of participants in movement activities? Does the strong impact of activism found among New Left activists in the United States hold in another country and across generations? To answer these questions, we test the relationship between engagement in social movement activities, namely participating in mass demonstrations, on the one hand, and a number of indicators of political life-course patterns in terms of political attitudes and behaviors, on the other. Our study has three main features, which, combined, we believe can lead us beyond existing scholarship on this topic. First, unlike most previous works in this field, we not only look at the biographical impact of participation by strongly committed activists of the New Left, but we examine whether run-of-the-mill activism, such as participating in mass demonstrations, has an effect on the political life of participants at the general population level as well. Second, unlike most of the existing studies, we conduct an analysis of such effects in a European country, namely Switzerland. Third, avoiding issues with retrospective data, we use a panel survey to disentangle the relationships between participation in demonstrations and in social movements more generally, on the one hand, and its individual-level effects, on the other. Panel data allow us to C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 87 [85–105] 15.9.2015 1:59PM 87 / The biographical impact of participation in social movement activities make a stronger case for a causal relationship between movement participation and its effects, since we measure the outcomes at a later time point than participation and control for the outcome variables at the time of participation also. By following over time the personal lives of people involved in social movement activities we can make a stronger case for the causal effect of participation in movements on the political life of individuals. The biographical impact of activism: cornerstones, blind spots, and promises Scholarly work on the biographical consequences of involvement in social movements has most often focused on New Left activists in the United States. A number of researchers have independently studied the long-term impact that involvement in protest activities has on both the attitudes and behaviors of those involved. The first major follow-up study of New Left activists was done by Demerath et al. (1971) in 1969, when they interviewed part of the volunteers they had surveyed four year earlier before and after the latter took part in a voter registration effort sponsored by the Southern Christian Leadership Conference. Some of the volunteers were then surveyed once again years later in order to gauge the long-term effects of their participation (Marwell et al. 1987). Similarly, Fendrich and his collaborators have studied a sample of activists involved in the civil rights movement in Tallahassee, Florida, in the early 1960s to show the biographical impact of activism (Fendrich 1974, 1977; Fendrich and Tarleau 1973). Here too, the subjects were surveyed once again at a later stage in order to assess the impact of their involvement in the long run (Fendrich 1993; Fendrich and Lovoy 1988). Other important studies of former New Left activists were conducted by Jennings and Niemi (1981), Nassi and Abramowitz (1979; see further Abramowitz and Nassi 1981), and Whalen and Flacks (1980, 1984, 1989). One of the most well-known works in this field is McAdam’s (1988, 1989) study of participants in the 1964 Mississippi Freedom Summer project. McAdam’s study was aimed to assess both the shortterm and long-term political and personal consequences of movement participation. Most importantly for our present purpose, it compared participants with a sample of “no-shows,” that is, individuals who applied, were accepted, but did not take part in the project. As a result, C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 88 [85–105] 15.9.2015 1:59PM 88 / Marco Giugni and Maria T. Grasso the researcher had at his disposal a control group of non-participants against which to assess the impact of participation. He found a strong and enduring effect of engagement on the political life of the participants. As summarized elsewhere (Giugni 2004; McAdam 1989), these follow-up studies of New Left activists provide consistent evidence that prior involvement in social movements and protest activities has a powerful and enduring impact of participation in movement activities on the biographies of participants. Specifically, activism had a strong effect both on the political and personal lives of the subjects. On the political side, former activists: (1) had continued to espouse leftist political attitudes (e.g., Demerath et al. 1971; Fendrich and Tarleau 1973; Marwell et al. 1987; McAdam 1989; Whalen and Flacks 1980); (2) had continued to define themselves as “liberal” or “radical” in political orientation (e.g., Fendrich and Tarleau 1973); and (3) had remained active in contemporary movements or other forms of political activity (e.g., Fendrich and Krauss 1978; Fendrich and Lovoy 1988; Jennings and Niemi 1981; McAdam 1989). On the personal side, former activists: (1) had been concentrated in teaching or other “helping” professions (e.g., Fendrich 1974; Maidenberg and Mayer 1970; McAdam 1989); (2) had lower incomes than their age peers; (3) were more likely than their age peers to have divorced, married later, or remained single (e.g., McAdam 1988, 1989); and (4) were more likely than their age peers to have experienced an episodic or nontraditional work history (e.g., McAdam 1988, 1989). While this body of scholarly work forms an invaluable cornerstone for research on the biographical consequences of participation in social movements, not least because of the consistency of findings across studies, there remain some blind spots. A number of such shortcomings are worth mentioning here (see Giugni 2004 for a more detailed discussion), as our study aims to go beyond them. In a way, they all refer to methodological issues in a broad sense. However, while some of them deal with the substantive focus of the study, others concern the methodological design of the research. On the substantive level, first, they all deal with a specific kind of movement participants, namely strongly committed activists, sometimes involved in high-risk political activities. As a result, they can say nothing about the personal consequences of more “routine,” low-risk forms of participation. Second, they study activists from a specific side of the political spectrum, namely New C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 89 [85–105] 15.9.2015 1:59PM 89 / The biographical impact of participation in social movement activities Left activists. Little can be said about participants in other movements. Third, they focus on a given historical period, namely the 1960s cycle of contention. This makes it hard to determine to what extent the characteristics in life-course observed in follow-up interviews are due to individual involvement in political activities rather than a product of the special era that forms the background of the research. Fourth, they draw their subjects from a specific national situation, namely that of the United States, and often from even narrower geographical areas. We do not whether the patterns found for activists in this country and these areas hold also in other places. In addition to the shortcomings related to the narrow focus, which undermines the representativeness of the samples, these studies suffer from some methodological problems more strictly defined, which prevent them from establishing a causal relationship between activism and its biographical effects. In particular, these studies often rely on retrospective data (but see the studies by Demerath et al., Jennings and Niemi, Marwell et al., and McAdam), with all the problems this entails in terms of potential biases when one relies on people’s recollection of previous attitudes or opinions, are done on non-representative samples, hence making empirical generalizations impossible, lack a control group (but see the studies by Fendrich, Fendrich and Lovoy, Jennings and Niemi, and McAdam), and are limited to a small number of subjects, going from a minimum of 11 in Whalen and Flacks’ study to a maximum of 330 in McAdam’s research. Overall, all these shortcomings weaken the follow-up studies of New Left activists to a considerable extent, in particular with regard to the cause–effect nexus and the generalization of empirical findings. More recently, scholars have addressed this topic on other kinds of movement participants and with new methods. In doing so, they have tried to address one or the other of the shortcomings of previous works, offering some important promises for further research. For example, Klatch (1999) studied the long-standing personal and biographical consequences of both people “on the left” and people “on the right” of the political spectrum. Taylor and Raeburn (1995)looked at the career consequences of high-risk activism by lesbian, gay, and bisexual sociologists. The most relevant work for our present purpose has been done by Sherkat and Blocker (1997). This study takes us beyond highly C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 90 [85–105] 15.9.2015 1:59PM 90 / Marco Giugni and Maria T. Grasso committed New Left activism to inquire into the individual-level effects of involvement in social movements by not-so-committed participants. Sherkat and Blocker (1997) investigated the political and personal consequences of participation in antiwar and student protests of the late 1960s using data from the Youth-Parent Socialization Panel Study. They found that ordinary involvement in these movements had an impact on the lives of those who had participated. They showed that demonstrators differed from non-activists in both their political orientations and certain life-course patterns, and that such differences were still present about a decade after their movement experiences. In particular, former protesters held more liberal political orientations and were more aligned with liberal parties and actions, selected occupations in the “new class,” were more educated, held less traditional religious orientations and were less attached to religious organizations, married later, and were less likely to have children. Our analyses follows Sherkat and Blocker’s (1997) footsteps in aiming to ascertain the effects of participation in mass demonstrations on the political-life of participants, but we do so in a context other than the United States. Compared to their study, we also look at how participating in demonstrations and more generally joining a social movement has an impact on individuals’ subsequent political life. We contribute to the literature in several important ways. First, we show that this effect holds also for participants others than those belonging to the New Left of the 1960s and 1970s. Second, we show it is true also of another country in addition to the United States, namely Switzerland. Third, we show this effect with time-ordered data measuring participation in demonstrations in 1999 and expected attitudinal and behavioral outcomes fifteen years later, in 2013, thus avoiding the problems with retrospective data. Fourth, we show that this effect is net of the previous state of such outcomes. In brief, our study draws on some of the advances made by these works in order to examine the potential impact of participation in social movements by not-so-committed participants who do not necessarily belong to the New Left in a country other than the United States. To do so, we make use of panel survey data. These data have two crucial features for our study. First, they are survey data, that is, they consist in a representative sample of the general population. As a result, they include both people who have participated in social movement activities as well as people who have not. Second, they are panel data, that is, they C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 91 [85–105] 15.9.2015 1:59PM 91 / The biographical impact of participation in social movement activities follow the same subjects over time. This allows us to make a stronger case for a causal effect of participation on the political life of participants. In addition, we are using panel survey data bearing on a country other than the United States, namely Switzerland. Is the consistently strong and durable impact of activism on the political life of New Left activists in the United States found in previous research also observed among run-of-the-mill movement participants in Switzerland? This is the question we will try to answer below. Before doing this, however, we need to discuss in more detail our data and operationalization. Data and methods To test our hypothesis that not only strong activism by a specific cohort in a specific country, but also less committed participation in social movements by a wider range of cohorts elsewhere has an impact on the subsequent political life or participants, we use data from the Swiss Household Panel. This is a yearly panel study of a random sample of private households in Switzerland. While its main goal is to observe social change, in particular the dynamics of changing living conditions and representations in the Swiss population, it also includes a number of indicators of political attitudes and behaviors, including participation in mass demonstrations. This panel has started with the first wave in 1999, the most recent available wave being that of 2013. We limited that sample to those born between 1934 and 1984, so that respondents in 2013 would have been younger than eighty and respondents in 1999 would have been older than fifteen, to allow for opportunities to participate in demonstrations without restriction of too young age of mobility. In the 1999 wave, respondents were asked whether they engaged in a demonstration in the last twelve months. We use two panel waves: 1999 and 2013. The 1999 waves provide the measure of movement participation we wish to assess the impact of as well as an initial measurement of our dependent variable that we can use as a “lagged” control. The 2013 wave allows us to ascertain whether participating in demonstrations has an effect on the subsequent political life of participants, as compared to non-participants, and gives us a sense of how durable such an effect (if any) is. By testing for politicallife outcomes of protesters fifteen years later, we can see whether even C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 92 [85–105] 15.9.2015 1:59PM 92 / Marco Giugni and Maria T. Grasso run-of-the-mill activism has an impact on the subsequent political life of individuals and whether it is an enduring impact. While these data are very useful in that they provide panel observations for a country other than the United States, they have several limitations that need to be mentioned. The most important one is that the question about participation in demonstration was asked only in the first wave. As a result, we do not know whether respondents who said in 1999 that they had participated in the last twelve month continued to attend afterwards, so that we cannot exclude that the observed effect is reinforced by protest in later years. Therefore, it may be that protesting in 1999 leads individuals to continue protesting and that this “protest life” leads to differences in values. Ideally, we would have liked to use methods for panel data analysis involving measures of all variables in all waves. However, the time-ordering of the data is still a large improvement on cross-sectional methods of analysis since it allows us to follow respondents over time and therefore avoid relying on retrospective data. In addition, it is possible to include measures of the dependent variables as observed in the first wave, so as to be able to examine effects net of previous levels. Hence, in each model we control for a measure of the dependent variable in 1999, whereas the latter is measured in 2013.1 This allows us to see whether demonstration participation in a previous wave distinguishes individuals in terms of their political-life outcomes in a subsequent wave, net of where they were at the previous point in time. This method is called the lagged dependent variable measure (Johnson 2005). In order to counter the risks of finding spurious relationships, in addition to including a lagged dependent variable in the models, we conducted two types of sensitivity analysis to provide supporting evidence for our conclusions from the main analysis. First, we also ran a model for each dependent variable in which we included all five lagged dependent variables. This test allows us to further control whether we see the effect of demonstrating due to some underlying unobserved variable common to both participants and non-participants. Second, we ran models in which we included an interaction term between attending demonstrations in 1999 and the lagged dependent variable 1 It should be noted that the membership in environmental organizations and party membership are measured in 2011 instead of 2013 since these questions were not asked in the 2013 nor in the 2012 panel waves. C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 93 [85–105] 15.9.2015 1:59PM 93 / The biographical impact of participation in social movement activities at hand also measured in 1999. This test allows us to control whether the effect of demonstrating is the same across values of the lagged dependent variables. The variables we use in our analyses are the following. Concerning the main independent variable, we coded engagement in protest in the last twelve months as 1 and other responses as 0. The specific indicator we use is participation in demonstrations. This is only one among a broader range of protest activities in which one may be involved. However, it is perhaps the most typical form, one that has become “modular” over time (Tarrow 2011), that is, used by a variety of actors for a variety of aims. When some have spoken of the “institutionalization of protest” (van Aelst and Walgrave 2001) or of the “social movement society” (Meyer and Tarrow 1998), they indeed refer to the widespread use of demonstrations as the archetypical social movement activity nowadays. Thus, while we cannot generalize beyond this specific form without testing empirically the effect of participating in other forms, this is a form that captures much of the social movements’ activities and political engagement therein. As to our dependent variable, we focus on five potential outcomes, covering both attitudinal (the first two) and behavioral (the last three) aspects. First, we look at the impact of participation in demonstrations on political interest. The latter is measured through a 0–10 scale, where 0 is lowest and 10 highest interest. Second, we have an indicator of self-placement on the left–right scale, where 0 is the Left and 10 the Right. This is meant to ascertain whether those who have demonstrated are more likely to change (or radicalize) their political ideology. Third, we consider voting behavior, more specifically voting for the left, as a first behavioral indicator. This is a dummy variable, whereby those who have voted for the left are coded 1, while those who have not are coded 0. We focus on left voting as movement participants are often characterized in the extant literature as being closer to the left (Dalton et al. 2010) Fourth, we examine membership in environmental organizations to wee whether movement participation impacts on organizational membership. This is a dummy variable, whereby members are coded 1, while non-members are coded 0. Fifth, we have a measure of party membership, so as to include institutional, in addition to non-institutional organizational involvement. Again, this is a dummy variable, whereby members are coded 1 and non-members 0. C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 94 [85–105] 15.9.2015 1:59PM 94 / Marco Giugni and Maria T. Grasso The regression analyses below include a number of controls relating to sociodemographic characteristics of respondents and their value orientations as identified as important in the literature on political participation: generation (WWII, baby boomers, and post-baby boomers), gender, education (degree), and a socioeconomic class (having a manual job) as sociodemographic controls. We also control for biographical availability, using marital status as a proxy (code 1 for being married, 2 for cohabiting, and 3 for being single). Controlling for these aspects is crucial to our test as they might affect the relationship between the fact of having demonstrated in 1999 and the political-life outcomes fifteen years later, in 2013. Findings Figure 4.1 shows our main findings. We present the results in graphical form in order to make them as readable as possible also to those not familiar with quantitative and especially regression techniques. We show five graphs, one for each dependent variable, or political life-course outcome.2 The graphs plot markers for coefficients and horizontal spikes for confidence intervals. The point estimates are unstandardized regression coefficients or log odds (depending on the regression model) based on lagged dependent variable models. When the segment indicating the confidence interval does not overlap with the vertical line, this means that the variable at hand is significantly different from zero, that is, the variable has a statistically significant effect. Otherwise, it has no significant effect. Of course, we are most interested in the variable concerning attendance of demonstrations in 1999. As we can see, demonstrating in 1999 has a significant effect on four out of five indicators of political-life outcome in 2013, net of their levels in 1999. These are: left–right self-placement, left voting, membership in environmental organizations, and party membership. For all four variables the horizontal spikes representing the standard errors do not overlap with the vertical line representing the null effect. Thus, people who said in 1999 that they had demonstrated in the twelve months prior to the interview were more likely than non-participants to vote for the left and to be members of an environmental organization 2 Table 4.1.A in the appendix shows the full results of the regression analyses. Left-right placement 2013 Left vote 1999 Left-right placement 1999 Demonstrated in 1999 Demonstrated in 1999 Born 1946/1964 (ref. 1934/1945) Born 1946/1964 (ref. 1934/1945) Born 1965/1984 (ref. 1934/1945) Born 1965/1984 (ref. 1934/1945) Born 1965/1984 (ref. 1934/1945) Gender (male) Gender (male) Gender (male) Cohabiting 2013 (ref. married) Cohabiting 2013 (ref. married) Cohabiting 2013 (ref. married) Single 2013 (ref. married) Single 2013 (ref. married) Single 2013 (ref. married) Degree 2013 Degree 2013 Degree 2013 Manual job 2013 Manual job 2013 Manual job 2013 0 .5 1 –1 Coefficient 0 1 2 –1 3 Member party 2011 Member environment org. 2011 Member party 1999 Member environment org. 1999 Demonstrated in 1999 Demonstrated in 1999 Born 1946/1964 (ref. 1934/1945) Born 1946/1964 (ref. 1934/1945) Born 1965/1984 (ref. 1934/1945) Born 1965/1984 (ref. 1934/1945) Gender (male) Gender (male) Cohabiting 2011 (ref. married) Cohabiting 2011 (ref. married) Single 2011 (ref. married) Single 2011 (ref. married) Degree 2011 Degree 2011 Manual job 2011 Manual job 2011 –1 0 1 Log odds 2 3 –.5 0 Coefficient Log odds –1 0 1 2 Log odds 3 4 Figure 4.1 Effect of participation in demonstrations in 1999 on selected indicators of political-life outcomes in 2013. .5 1 [85–105] 15.9.2015 1:59PM –.5 95 Demonstrated in 1999 Born 1946/1964 (ref. 1934/1945) –1 C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D Left vote 2013 Political interest 2013 Political interest 1999 C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 96 [85–105] 15.9.2015 1:59PM 96 / Marco Giugni and Maria T. Grasso in 2103. In addition, they were also more likely than non-participants to be members of a political party, attesting to the linkage between institutional and contentious politics. More than that: they also tended to become more leftist in their political orientations, at least based on their self-placement on the left–right scale. The negative sign of the coefficient suggests that participating in demonstration makes people more leftist in their political orientation, as 0 means left and 10 means right on this scale. This is in our view a particularly interesting finding. While the social movement literature argues that those attending demonstrations in general tend to be on the left-side end of the political spectrum, here we show that the very fact of participating makes them become even more leftist, providing a reinforcing effect for further movement participation. While both attitudinal and behavioral measures of political-life outcomes are significantly affected by previous participation in demonstrations, we observe no effect on political interest. In other words, those who participated in demonstrations in 1999 are not more likely to see an increase in their political interest between 1999 and 2013 relative to those who did not participate. However, the significant effect of party membership in 1999 on party membership in 2013 suggests that a certain outcome at time T0 has a strong effect on the same outcome at time T1. In other words, those who were interested in politics remained so fifteen years later. The same applies to all four other dependent variables. From a more substantive point of view, there is quite a consistent effect of gender. Specifically, men are more interested in politics, place themselves more on the right of the political spectrum, are more likely to be members of a political party, and less likely to vote for the left. The only variable that is not affected by gender is membership in environmental organizations. The effect for the baby boomer generation – that is, the cohort that was active in the New Left – is not significant, suggesting that at least in Switzerland this cohort was not particularly left-wing or active net of the other effects we control for. We also observe an effect of the other three controls, namely marital status, education, and social class, albeit not always a consistent one across all five models. Cohabiting respondents are more likely than married ones (the reference category for this variable) to vote for the left. This is the only statistically significant effect concerning marital status. Consistently with the literature on political participation (Brady et al. C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 97 [85–105] 15.9.2015 1:59PM 97 / The biographical impact of participation in social movement activities 1995; Verba et al. 1995), education has a positive impact on all the dependent variables, that is, on all the measures of political engagement. Finally, social class plays a role as well, as respondents who have a manual job are more likely to vote for the left and to be members of an environmental organization, as compared to other professions. Overall, these effects are in line with the mainstream literature on political participation and more specifically individual engagement in social movements. We thus find an impact of participation in demonstrations on a number of political attitudes and behaviors fifteen years later. As we said in the methodological section, in order to check for the robustness of our results, we performed two additional sensitivity analyses. First, we ran the same models but with the inclusion of all lagged dependent variables in each model. So, for example, the model predicting political interest in 2013, in addition to political interest in 1999, also includes self-placement on the left–right scale, voting for the left. Membership in environmental organizations, and party membership, all measured in 1999. The results (not shown) point in the same direction as the main analysis: three out of four indicators of political-life outcomes – selfplacement on the left–right scale, voting for the left, and membership in environmental organizations – remain statistically significant below the standard 5% threshold, while the fourth one – party membership – is significant but only at the 10% level. Second, we ran models including an interaction term between attending demonstrations in 1999 and the lagged dependent variable. Again, the results (not shown) of this further check underlines the robustness of our results, as none of the interactive terms is statistically significant. This means that the effect of demonstrating was the same across values of our lagged dependent variables. Finally, we also ran the same models as in our main tests, but on the dependent variables measured one year later only, in 2000. If we found the effect of attending demonstrations in 1999 on our five dependent variables one year later to be smaller than those observed for 2013 or even inexistent, we can conclude, at least tentatively, that the impact of participation in demonstrations is a long term one. Under control of the lagged dependent variable, the coefficients are all smaller than in the analysis for the 1999–2013 time span, and the effect for one of the dependent variables – party membership – disappears. Furthermore, when controlling for all lagged dependent variable in each model, only C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 98 [85–105] 15.9.2015 1:59PM 98 / Marco Giugni and Maria T. Grasso the effect for self-placement on the left–right scale remain significant, while the one for voting for the left remain significant but only at the 10% level. This provides further evidence of a long-term impact of attending demonstrations. Overall, our findings confirm what found by previous research on the biographical consequences of activism in the United States, hence strengthening them to the extent that one can start generalizing beyond the specific national case. They also confirm in particular the findings of Sherkat and Blocker (1997). They suggest that previous participation in mass demonstrations has a significant and enduring impact on the political life of participants. In other words, even low-risk activism distinguishes people on political-life outcomes of movement participation. Most importantly, these effects are durable, as they are observed after fifteen years. Conclusion This chapter has explored the question of the biographical outcomes of activism. In particular, it employed panel data from a general population household survey from a country other than the United States to examine whether participation in protest activities has an impact on the subsequent political life of participants. Results showed that it does. Specifically, previous participation in demonstrations in Switzerland significantly affected participants’ political attitudes and behaviors when they were interviewed a year later. Most importantly, we found that impact to be visible later in time, namely fifteen years later. Thus, according to our analysis, demonstrating has an important and durable effect on such political-life outcomes as self-placement on the left–right scale, voting for the left, membership in environmental organizations, and party membership. The only indicator for which we found no evidence is political interest. We have understood attending a demonstration as a marker of deeper political involvement with social movements. Our analysis suggests that participating in demonstrations and, more generally, being active in social movements changes people and has an impact on their subsequent political life, both at the attitudinal and the behavioral level. We have shown that this effect is net of their previous level of political engagement as well as of their current social C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 99 [85–105] 15.9.2015 1:59PM 99 / The biographical impact of participation in social movement activities characteristics and biographical availability. It stands to reason that joining a social movement would lead individuals to keep their commitment to contentious politics, but also to become engaged in more institutional forms of participation. Furthermore, it also seems to lead them to become more leftist in their political orientations, be it in terms of self-placement on the left–right scale or in terms of voting for the left. What lessons can we draw from our findings and what avenues for future research do they open up? The first and foremost lesson is that the biographical impact of participation in social movement activities is not limited to highly committed New Leftactivists. As was discussed in the literature review, previous research has abundantly shown that activism by American New Left in the late 1960s and early 1970s had a strong effect both on the political and personal lives of participants (see Giugni 2004 for a review). Our analysis suggests that not only strongly committed activists, but also run-of-the-mill activists undergo a transformation through participating in social movements. Their political attitudes and behavioral patterns in their later lives are changed through this earlier involvement. Thus, we have controunuted to the still rare scholarship on individual-level effects of involvement in social movements by not-so-committed participants (McAdam 1999a; Sherkat and Blocker 1997; Van Dyke et al. 2000; Wilhelm 1998). These studies have used random samples of American citizens to examine the impact of ordinary involvement in social movements on the lives of those who had participated, showing that demonstrators differed from non-activists in both their political orientations and certain lifecourse patterns. We have replicated this research on a different sample using the Swiss Household Panel and obtained similar findings as far as certain political attitudes and behaviors are concerned. The timeordered nature of our data avoided issues with retrospective data and problems of recall. The importance of our findings lies above all in the fact that they allow scholarship to go beyond highly committed New Left activism. Although our analysis should be developed in future research, both theoretically and methodologically, our results suggest that the biographical outcomes of social movements are not confined to a specific generation of strongly committed activists mobilizing during the 1960s and 1970s within the New Left in the United States. Rather, our results show that these findings can be generalized to other cohorts C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 100 [85–105] 15.9.2015 1:59PM 100 / Marco Giugni and Maria T. Grasso of people involved in social movements activities with much lower intensity and in different countries, at a different point in time. The fact that we used survey data to generalize existing findings to a new country, other cohorts, and participants in general population samples allows us to draw wider conclusions about whether participation in social movement activities can translate into aggregate-level outcomes, in other words, whether protest participation can be seen to alter the political life of the population more generally and not just those of a handful of highly committed activists living in self-consciously counter-cultural enclaves. As was shown by McAdam (1999a), there is a broader life-course impact of movement activity. If confirmed by further research, our findings are important for the study of the consequences of social movements. They show that attending a demonstration can change people. Participation in social movement activities influences individuals in certain key aspects of their political life such as their ideology or positioning on the left–right scale, their voting behavior, and also their organizational membership. While such biographical outcomes are readily understandable when it comes to strongly committed social movement activists who dedecate their whole lives, sometimes paying a high price, for the movement, how is it that the simple act of participating in a demonstration can change attitudes for the long-term even amongst the general population? The answer to this question calls for a reflection on the mechanisms linking participation in demonstrations to the political life of participants. One possible explanation is that attending mass demonstrations works as an initiator of “cognitive liberation” (McAdam 1999b): it allors for individuals to realize the importance of political engagement and action to improve the world we live in, not only through continued activism, but also in joining a social movement organization or, on the institutional side, a political party. The effect may also be more “structural”: attending a demonstration also allows for meeting other, politically engaged, people. So, just as joining a social movement can be explained, at least in part, by pre-existing social ties and interpersonal networks (McAdam and Paulsen 1993; Passy 2003), the very fact of participating puts people in contact with similarly minded individuals, therefore providing a reinforcinf mechanism stimulating the deepening of one’s political beliefs. Exchanges and communications occurring during or in the aftermath of attending a demonstration would contribute to making the participant more politically aware and committed. C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 101 [85–105] 15.9.2015 1:59PM 101 / The biographical impact of participation in social movement activities Moreover, this process would also contribute to the formation of a collective identity, crucial in particular for engagement in social movements (Polletta 2001). Additionally, it may be that participation reinforces attitudes that lead to repeated participation. In this way, attendance at a demonstration can initiate one to a new “protest life”’ that leads to further, more wide-ranging personal changes. Further research is needed in order to test the internal and external validity of our findings. Concerning the former, as was stressed in the methodological section, panel data with measures available at all points in time, would be desirable. Concerning the latter, for example, future research should aim to test whether we observe similar patterns in yet other contexts.. Regardless, our findings show that something about participating to protest activities has important effects on political lifecourse patterns. This could be since individuals engaged in social movement activities have the courage, the mental structures, the resources, and so forth to remain committed to social and political causes or to become even more committed to them. Future research should aim to test and verify these mechanisms. Appendix Table 4.1.A Lagged dependent variable regression models (1999–2011/2013) Membership environmental organization 2011 Political interest 2013 Left–right selfplacement 2013 0.04 (0.21) −0.78*** (0.19) 1.22*** (0.30) 0.97*** (0.25) 0.66* (0.31) Born 1946/1964 −0.37 (ref.: 1934/1945) (0.24) 0.28 (0.20) −0.10 (0.35) 0.08 (0.34) −0.16 (0.38) Born 1965/1984 −0.64** (ref.: 1934/1945) (0.25) 0.40 (0.21) −0.30 (0.36) 0.16 (0.35) −0.26 (0.40) Male 0.40*** (0.09) −0.38* (0.15) 0.02 (0.13) 0.51** (0.18) Demonstrating 1999 0.46*** (0.10) Left voting 2013 Party membership 2011 C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 102 [85–105] 15.9.2015 1:59PM 102 / Marco Giugni and Maria T. Grasso Table 4.1.A Cont. Membership environmental organization 2011 Political interest 2013 Left–right selfplacement 2013 Cohabiting (ref.: married) 0.32 (0.19) −0.01 (0.17) 0.74** (0.28) 0.26 (0.25) −0.14 (0.38) Single (ref.: married) 0.07 (0.16) −0.18 (0.14) 0.19 (0.24) 0.38 (0.19) −0.31 (0.30) Education (degree) 0.37** (0.12) −0.27** (0.10) 0.37* (0.17) 0.40** (0.16) 0.52** (0.20) −0.21 (0.14) 0.09 (0.13) −0.52* (0.25) −0.91*** (0.23) −0.44 (0.27) Lagged dependent variable (1999) 0.59*** 0.51*** 2.64*** 2.50*** 3.14*** (0.02) (0.02) (0.15) (0.14) (0.19) Constant 2.74*** (0.26) 2.16*** (0.23) −1.90*** −2.21*** (0.36) (0.35) −2.85*** (0.40) N R-squared / Pseudo R-squared Log likelihood 1569 0.472 1307 0.372 1324 0.294 1797 0.238 1796 0.281 −563.61 −739.90 −487.32 Manual job −3235.74 −2437.19 Left voting 2013 Party membership 2011 * p < 0.05, ** p < 0.01, *** p < 0.001 Notes: OLS models for political interest and left–right self-placement. Logit models for leftvoting, membership in environmental organizations, and party membership. Standard errors in parentheses. The lagged dependent variables are the same as the dependent variable in each model but measured in 1999, at the same time as demonstrating. Controls are from the same survey year as the dependent variable. Since, there was no survey question for associational membership in 2013, nor in 2012, the measures of membership in environmental organizations and party membership are from 2011. References Abramowitz, Stephen I. and Alberta J. Nassi. 1981 “Keeping the Faith: Psychological Correlates of Activism Persistence into Middle Adulthood.” Journal of Youth and Adolescence, 10: 507–523. Brady, Henry E., Sidney Verba, and Kay Lehman Scholzman. 1995. “Beyond SES? A Resource Model of Political Participation.” American Political Science Review, 89: 271–294. C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 103 [85–105] 15.9.2015 1:59PM 103 / The biographical impact of participation in social movement activities Dalton, Russell, Alix van Sickle, and Steven Weldon. 2010. “The Individual-Institutional Nexus of Protest Behaviour.” British Journal of Political Science, 40: 51–73. Demerath, N.J., Gerald Marwell and Michael Aiken. 1971. Dynamics of Idealism. San Francisco: Jossey-Bass. Fendrich, James M. 1974. “Activists Ten Years Later: A Test of Generational Unit Continuity.” Journal of Social Issues, 30: 95–118. Fendrich, James M. 1977. “Keeping the Faith or Pursuing the Good Life: A Study of the Consequences of Participation in the Civil Rights Movement.” American Sociological Review, 42: 144–157. Fendrich, James M. 1993. Ideal Citizens. Albany: State University of New York Press. Fendrich, James M. and A.T. Tarleau. 1973. “Marching to a Different Drummer: Occupational and Political Correlates of Former Student Activists.” Social Forces, 52: 245–253. Fendrich, James M. and Elis M. Krauss. 1978. “Student Activism and Adult Left-Wing Politics: A Causal Model of Political Socialization for Black, White and Japanese Students of the 1960s Generation.” Research in Social Movements, Conflict and Change, 1: 231–256. Fendrich, James M. and Kenneth L. Lovoy. 1988. “Back to the Future: Adult Political Behavior of Former Political Activists.” American Sociological Review, 53: 780–784. Giugni, Marco. 2004. “Personal and Biographical Consequences.” In David A. Snow, Sarah Soule, and Hanspeter Kriesi (eds.), The Blackwell Companion to Social Movements. Oxford: Blackwell, 489–507. Jennings, M. Kent and Richard G. Niemi. 1981. Generations and Politics. Princeton, NJ: Princeton University Press. Johnson, David. 2005. “Two-Wave Panel Analysis: Comparing Statistical Methods for Studying the Effects of Transition.” Journal of Marriage and Family, 67: 1061–1075. Klatch, Rebecca. 1999. A Generation Divided. Berkeley: University of California Press. Maidenberg, Michael and Philip Meyer. 1970. “The Berkeley Rebels Five Years Later: Has Age Mellowed the Pioneer Radicals?” Detroit Free Press, February 1–7. Marwell, Gerald, Michael Aiken, and N.J. Demerath. 1987. “The Persistence of Political Attitudes among 1960s Civil Rights Activists.” Public Opinion Quarterly, 51: 359–375. C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 104 [85–105] 15.9.2015 1:59PM 104 / Marco Giugni and Maria T. Grasso McAdam, Doug. 1988. Freedom Summer. New York: Oxford University Press. McAdam, Doug. 1989. “The Biographical Consequences of Activism.” American Sociological Review, 54: 744–760. McAdam, Doug. 1999a. “The Biographical Impact of Activism.” In M. Giugni, D. McAdam, and C. Tilly (eds.), How Social Movements Matter. Minneapolis: University of Minnesota Press, 117–146. McAdam, Doug. 1999b [1982]. Political Process and the Development of Black Insurgency, 1930–1970, Second edition. Chicago: University of Chicago Press. McAdam, Doug and Ronnelle Paulsen. 1993. “Specifying the Relationship Between Social Ties and Activism.” American Journal of Sociology, 99: 640–667. Meyer, David S. and Sidney Tarrow (eds.). 1998. The Social Movement Society: Contentious Politics for a New Century. Lanham, MD: Rowman and Littlefield. Nassi, Alberta J. and Stephen I. Abramowitz. 1979 “Transition or Transformation? Personal and Political Development of Former Berkeley Free Speech Movement Activists.” Journal of Youth and Adolescence, 8: 21–35. Passy, Florence. 2003. “Social Networks Matter. But How?” In Mario Diani and Doug McAdam (eds.), Social Movements and Networks: Relational Approaches to Collective Action. Oxford: Oxford University Press, 21–48. Polletta, Francesca and James M. Jasper. 2001. “Collective Identity and Social Movements.” Annual Review of Sociology, 27: 283–305. Sherkat, Darren E. and Blocker T. Jean. 1997: Explaining the Political and Personal Consequences of Protest. Social Forces, 75, 1049–1070. Taylor, Verta and N. C. Raeburn. 1995. “Identity Politics as High-Risk Activism: Career Consequences for Lesbian, Gay, and Bisexual Sociologists.” Social Problems, 42: 252–273. Tarrow, Sidney. 2011. Power in Movement: Social Movements and Contentious Politics. Cambridge: Cambridge University Press. Van Aelst, Peter and Stefaan Walgrave. 2001. “Who is that (Wo)man in the Street? From the Normalisation of Protest to the Normalisation C:/ITOOLS/WMS/CUP-NEW/6684132/WORKINGFOLDER/BOSI/9781107116801C04.3D 105 [85–105] 15.9.2015 1:59PM 105 / The biographical impact of participation in social movement activities of the Protester.” European Journal of Political Research, 39: 461–486. Van Dyke, Nella, Doug McAdam, and Brenda Wilhelm. 2000. “Gendered Outcomes: Gender Differences in the Biographical Consequences of Activism.” Mobilization, 5: 161–177. Verba, Sidney, Kay Lehman Schlozman, and Henry E. Brady. 1995. Voice and Equality: Civic Voluntarism in American Politics. Cambridge, MA: Harvard University Press. Whalen, Jack and Richard Flacks. 1980. “The Isla Vista ‘Bank Burners’ Ten Years Later: Notes on the Fate of Student Activists.” Sociological Focus, 13: 215–236. Whalen, Jack and Richard Flacks. 1984 “Echoes of Rebellion: The Liberated Generation Grows Up.” Journal of Political and Military Sociology, 12: 61–78. Whalen, Jack and Richard Flacks. 1989. Beyond the Barricades. Philadelphia: Temple University Press. Wilhelm, Brenda. 1998 “Changes in Cohabitation across Cohorts: The Influence of Political Activism.” Social Forces, 77: 289–310.