Unconventional protests: Partisans and independents outside the Republican and Democratic national conventions Research and Politics October-December 2016: 1 –8 © The Author(s) 2016 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/2053168016674138 rap.sagepub.com Creative Commons Non Commercial CC-BY-NC: This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 3.0 License (http://www.creativecommons.org/licenses/by-nc/3.0/) which permits non-commercial use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). Protests at national party conventions have long been a part of the political landscape in the United States. For example, on June 8, 1908, Samuel Gompers, President of the American Federation of Labor, led a march for labor rights outside the Republican National Convention (RNC) in Chicago (Staff, 1908). Over the years, conventions have seen protests for woman suffrage, civil rights, farm subsi- dies, peace in Vietnam and Iraq, the right to life and the right to choose, tighter financial regulations on banks, and many other issues. Activists travel to conventions to express their viewpoints because these gatherings symbolize the parties as institutions in a way that no other events do (Edelman, 1964: 95). Activists are often motivated to mobi- lize outside conventions because it is a relatively easy way to seize media coverage for their causes (Sobieraj, 2011). Convention protests have been especially contentious in the 2000s, which have seen frequent clashes between protest- ers and police, as well as mass arrests (Hermes, 2015). Despite the regular appearance of convention protests during the presidential election cycle, relatively little is known about why individuals participate in this type of activism. Are they there to help or hurt a candidate? To pro- mote an issue? To express themselves? To see friends? Or some mixture of all of these? What factors shape these motivations? This lacuna is unfortunate because protest organizers are among the policy demanders that work to shape party coalitions (Bawn et al., 2012). Through these protests, parties and movements meet in the streets outside the conventions in a tangible way. A sizeable body of research has emerged on how leaders in parties and move- ments influence one another (see, e.g., Clemens, 1997; Frymer, 1999; Schlozman, 2015). Yet, how parties and movements affect, and are affected by, rank-and-file activ- ists remains less-explored territory (but see Heaney and Unconventional protests: Partisans and independents outside the Republican and Democratic national conventions Michael T. Heaney Abstract Protests at national party conventions are an important setting in which political parties and social movements challenge one another. This article examines the motivations of participants in these events. Drawing upon data from surveys of protesters outside the 2008 national party conventions, it focuses on how partisan and independent political identifications correspond with the reasons that individuals give for protesting. The results demonstrate that there are some conditions under which independents place a greater focus on issues than do partisans and under which partisans place a greater focus on presidential candidates than do independents. However, there are also conditions under which independents are inclined to work alongside partisans, such as trying to stop the election of a threatening candidate and in championing an issue outside their opposing party’s convention. The article argues that micro-level partisan identifications are thus likely to affect the broader structure of party coalitions. These considerations promise to become increasingly relevant as social movements – such as the Tea Party, Occupy Wall Street, and Black Lives Matter – launch new campaigns against or within established parties. Keywords Political party, social movement, partisan identification, independent, Democratic National Convention, Republican National Convention Organizational Studies Program and Department of Political Science, University of Michigan, Ann Arbor, MI, USA Corresponding author: Michael T. Heaney, Organizational Studies Program and Department of Political Science, University of Michigan, 2016-J Ruthven Museums Building, 1109 Geddes Avenue, Ann Arbor, MI 48109-1079, USA. Email: mheaney@umich.edu 674138RAP0010.1177/2053168016674138Research & PoliticsHeaney research-article2016 Research Article mailto:mheaney@umich.edu http://crossmark.crossref.org/dialog/?doi=10.1177%2F2053168016674138&domain=pdf&date_stamp=2016-10-19 2 Research and Politics Rojas, 2007, 2015; Layman et al., 2010; Skocpol and Williamson, 2012). In examining this topic, Fabio Rojas and I (Heaney and Rojas, 2007, 2015) theorize that political parties and social movements sometimes come together in a social space that we call “the party in the street.” This space provides an opportunity for activists to pressure a party to pursue a movement’s issues and, reciprocally, to pressure the move- ment to support a party’s candidates. We argue that partisan identification draws activists into politics when it serves the party’s interests, but also pushes activists out of politics when their involvement no longer serves the party’s goals. Thus, we see a trade-off between party and movement activism – which is largely to the advantage of parties – at least during periods of high partisan polarization. However, this empirical analysis is limited by the fact that it focuses on issue involvement without considering the actions that activists may try to help or hurt candidates directly. Samara Klar and Yanna Krupnikov (2016) also examine how partisan identification affects political participation, though they stress the consequences of the decision by an increasing number of citizens to identify as independents, who avoid making public commitments to one of the major parties. They argue that even though independents may tend to vote as if they are loyal partisans, their embarrass- ment with the parties and with contemporary politics may motivate them to refrain from disclosing their partisan lean- ings. In particular, independents tend to avoid the kind of interpersonal activism that promotes the flow of partisan ideas through social networks, such as talking to friends about politics and displaying partisan bumper stickers, yard signs, and buttons. This avoidance of partisan engagement on the part of independents is an impediment to the parties’ efforts to win elections and govern. In contrast to Heaney and Rojas’s (2007, 2015) claim that parties gain from the participation of independents when they are allied with partisans, Klar and Krupnikov emphasize that parties lose when independent identifiers do not participate in politics. Yet, Klar and Krupnikov do not examine what happens when independents do participate in partisan politics. How is their participation different from that of party identifiers? Moreover, their empirical analysis focuses on actions that independents may or may not take to help candidates, but does not consider the potentially powerful actions that independents may take to hurt candidates. In this article, I argue that combining the Heaney–Rojas and Klar–Krupnikov perspectives on partisan and inde- pendent identifications provides significant insights into the competing motivations of protesters outside the national party conventions. It considers that activists may act to help or hurt a candidate, or speak out on an issue, within the same analysis. I argue that partisan identification – or its absence – affects the mixed conditions under which activ- ists seek to be a resource to parties or movements. The article proceeds, first, by discussing the role of par- tisan identification and independence in individuals’ par- ticipation in activism. Second, it outlines the research design based on a survey of demonstrators outside the 2008 national party conventions. Third, it presents descriptive survey results. Fourth, it reports estimates of probit models for protest motivation. It concludes by considering how political parties and social movements relate to one another at the grassroots level. Partisans, independents, and activism Political activism is a widely prevalent form of political participation, though it receives substantially less attention than does voting behavior. Activism reflects individuals’ involvement in politics that takes place informally, such as by participating in demonstrations, writing about issues online, organizing citizens’ groups, and writing letters to elected officials. Using data from a longitudinal study of activism from 1965 to 1997, Catherine Corrigall-Brown (2012) reports that approximately two-thirds of Americans engage in some form of political activism over their life- times. Schlozman et al. (2012) argue that this activism redirects the attention of politicians away from the prefer- ences of the median voter and toward the causes champi- oned by activists. I argue that partisan identification – or the lack thereof – is closely linked with the ways that individuals make sense of their participation in political activism. People that identify closely with a political party may turn to party poli- tics to understand and explain their involvement in activ- ism. They are more likely to be attracted to the competitive aspects of elections that are integral to party politics (Schattschneider, 1960). People who describe themselves as independents are likely to have different types of concerns about politics than do partisans. As Klar and Krupnikov (2016) point out, inde- pendents may be put off by the conflictual aspects of poli- tics to a degree that self-identified partisans may not be. However, they may also reject partisan identities because they are cross-pressured on issues with which they strongly identify (Magleby and Nelson, 2012). Social identities related to these issues, then, may help to motivate protest participation (Polletta and Jasper, 2001). Convention pro- tests usually have a strong issue orientation that is likely to amplify these motivations. As a result, the independents who protest at conventions may not be in the “middle of the road” between the Democratic and Republican parties. Rather, they may identify as independents because there is at least one issue on which they do not think that the parties are extreme enough. Building on these arguments, it is possible to state a series of hypotheses related to partisan identification, inde- pendence, and the motivation of individuals for participat- ing in activism outside conventions. First, partisans at Heaney 3 convention protests are likely to want to help the presiden- tial candidate of their party, especially if they are attending the convention at which that candidate is nominated. Thus: H1: Partisans are more likely than independents to be motivated by the desire to help a candidate if they are protesting at their more-proximate party’s convention. Second, it is possible that activists protest to hurt the chances of the candidate from the other party. They are more likely to see the opportunity to do so if they are attend- ing the convention of the opposing party. Thus: H2: Partisans are more likely than independents to be motivated by the desire to hurt a candidate if they are protesting at their more-distant party’s convention. Third, partisans and independents are likely to be differ- ently motivated by issues. Because their energy is drawn into the competitive aspects of partisan politics, partisans are less likely to articulate their motivations explicitly in issue terms. In contrast, independents are more likely to see issues as closer to their political identities. These propensi- ties underlie the trade-off between party and movement activism. Thus: H3: Partisans are less motivated to protest by issue con- siderations than are independents. Underlying these hypotheses is the more general hypothe- sis that independents are differently motivated than parti- sans in attending protests. Independents generally eschew working to affect the chances of candidates and instead concentrate on their issue concerns. Thus: H4: Partisans and independents are differently moti- vated to protest in that partisans place a greater empha- sis on candidates and independents place a greater emphasis on issues. Convention protest surveys This study is based on surveys conducted of protesters out- side the 2008 RNC in St Paul, Minnesota and 2008 Democratic National Convention (DNC) in Denver, Colorado. The attitudes of the participants in the protests were assessed using a five-page, pencil-and-paper survey.1 The questions queried individuals’ motivations for partici- pating in the protest, partisan identification, ideology, emo- tional engagement in politics, social networks with other protesters, where they live, sex/gender, race/ethnicity, age, level of education, and annual income, as well as other atti- tudinal and demographic questions. Surveys were collected at all protests using the “anchor sampling” method developed by Heaney and Rojas (2007, 2014, 2015) as an adaptation of exit-poll sampling. The sur- vey team began by distributing itself throughout the major regions of the crowd. Surveyors were instructed to look into the crowd and select a person to serve as an “anchor” for their counting process. This anchor could be anyone they preferred, though the anchor would not be surveyed on the understanding that s/he might be selected with bias. From the anchor, surveyors counted five persons to the right, and then issued a survey invitation; again, counted five to the right, then issued a survey invitation, continuing until three individuals had accepted the survey. Once three surveys were completed, the surveyor selected a new anchor and repeated the procedures. Non-responses were recorded for the purpose of tabulating response rates. Research shows that if these surveys are conducted with attention to ensuring random selection and account for biases in non-response, then they can generate samples that are a good representation of the individuals attending the protest (Walgrave and Verhulst, 2011). Who are convention protesters? The survey yielded 412 respondents from the DNC, with a response rate of 65 percent, and 990 respondents from the RNC, with a response rate of 77 percent. All of the analysis reported here is weighted according to differences in response rates and accounts for different variances in the two strata of the survey (i.e. the DNC and the RNC). The partisan breakdown of protest participants is pro- vided in Figure 1. The distribution has a strong, overall skew toward non-Republicans. Republicans and people who lean Republican made up about 6 percent of protesters at the RNC and 11 percent of protesters at the DNC. Among other participants, there was a greater percentage of inde- pendents at the DNC and a greater percentage of Democrats at the RNC. Roughly 36 percent of protesters at the RNC were independents, as were 44 percent of respondents at the DNC. Democrats and Democratic leaners made up 45 per- cent of protesters at the DNC and 57 percent at the RNC. Respondents were asked “Which are the most important reasons that you decided to come to this convention?” They were shown a list of 12 options (plus “other”) and asked to circle up to three of them. This article focuses on three of these options: (1) “To help a candidate win this year’s pres- idential election”; (2) “To help prevent a candidate from winning this year’s presidential election”; and (3) “To express my view on a particular issue or issues.” Respondents who indicated that they were motivated by issues were also asked to “Please specify the top issues.” Selected motivations are reported by convention in Figure 2.2 Respondents were more likely to seek to help a candidate at the DNC (14%) than at the RNC (10%) and more likely to seek to hurt the chances of a candidate at the RNC (18%) than at the DNC (7%). However, issues pro- vided a more powerful source of inspiration for these 4 Research and Politics respondents.3 Approximately 32 percent of DNC protesters and 44 percent of RNC protesters were motivated by issues.4 War and peace was the issue most commonly men- tioned, with 27 percent of respondents volunteering this motivation. There was a significant fall-off after war/peace, with 6 percent concerned about the economy, 5 percent mentioning health, 4 percent calling for group rights, 3 per- cent pointing to problems with the environment, and all other issues falling below this threshold. Models of protest motivation This article estimates three probit regression models of pro- test motivation: one for helping a candidate, one for hurting the chances of a candidate, and one for issue motivation. Complete-case imputation (Little, 1988) was used to replace the values of missing data, which is an appropriate method given the low percentage of missing data (King et al., 2001). Each equation includes the same independent variables. The first focus variable is More-Proximate Party Convention. This variable takes the value of 1 if a Strong Democrat, Not Very Strong Democrat, or Independent Who Leans Democrat is protesting the DNC; or, if a Strong Republican, Not Very Strong Republican, or Independent Who Leans Republican is protesting the RNC; 0 otherwise. The second focus variable is More-Distant Party Convention. This variable takes the value of 1 if a Strong Figure 1. Partisan identification of protesters by convention. Figure 2. Selected motivations of protesters by convention. Heaney 5 Democrat, Not Very Strong Democrat, or Independent Who Leans Democrat is protesting the RNC; or, if a Strong Republican, Not Very Strong Republican, or Independent Who Leans Republican is protesting the DNC; 0 otherwise. The implicit excluded category is independent, which com- bines both middle-independent category and the other- independent category. The remaining variables are included for the purpose of statistical control. Ideology is measured on a 9-point ordinal scale from left to right, with “To the ‘right’ of strong conservative” receiving a score of 9 and “To the ‘left’ of strong liberal” receiving a score of 1. Enthusiasm about Politics and Anxiety about Politics are included in order to account for positive and negative emotional engagement with the political process. They are measured on a 5-point ordinal scale, with people reporting enthusi- asm and anxiety “Almost Always” receiving a 5 and peo- ple who “Never” experience enthusiasm and anxiety receiving a 1. Social Networks at Protest accounts for the possibility that people that perceive that they are well con- nected at the protest may be motivated differently than people that perceive that they are poorly connected at the protest. Live in State of Convention controls for the pos- sibility that local people are differently motivated than those that travel a considerable difference to protest the convention. Sex/Gender takes the value of 1 for women and 0 for men. Race/Ethnicity takes the value of 1 for whites and 0 for nonwhites. Age is measured in years. Education is measured on a 6-point ordinal scale, with “Graduate or professional degree” receiving a score of 6 and “Less than high school diploma” receiving a score of 1. Income is measured in thousands of dollars for 2007. Probit results are reported in Table 1. The estimates in equation (1) are consistent with H1. They show that when partisans are protesting at their more-proximate party con- vention, they are more likely than independents to do so for the purpose of helping a candidate for president. Additionally, consistent with H4, the results show that par- tisans protesting at their more-distant party’s convention are also more likely than independents to want to help a candidate for president. That is, they may see their protest of the opposing party directly in terms of being helpful to their home party. On this point, independents are clearly different from partisans: they are very unlikely to protest with the intention of helping a candidate. Only about 6 per- cent of independents reported doing so. The estimates in equation (2) are consistent with H2. They show that when partisans are protesting at their more- distant party’s convention, they are more likely than inde- pendents to work to hurt the chances of a candidate. However, the results are not entirely consistent with H4. The estimates reveal that independents are significantly more likely than partisans attending their more-proximate party’s convention to protest with the intention to hurt the chances of a candidate. Thus, rather than being strictly less interested in candidate politics than are partisans, inde- pendents occupy a middle ground. About 10 percent of independents were motivated to stop a candidate, which was more than the 4 percent of partisans at their more- proximate convention and less than the 25 percent of parti- sans at their most-distant convention. The estimates in equation (3) support H3. They demon- strate that partisans at their more-proximate party conven- tion are less likely than independents to protest because of issue concerns. However, again, the results are not entirely consistent with H4. There is no statistically significant dif- ference in the degree to which they are motivated by issues between independents and partisans when they are attend- ing their more-distant party’s convention. About 40 percent of independents and 46 percent of partisans at their more- distant convention came out due to issues. In contrast, only 28 percent of partisans at their more-proximate convention mentioned issues as one of their reasons. The effects of the control variables exhibited variation across the equations. Anxiety about politics had significant, positive correlations across all three equations, though enthusiasm about politics only significantly contributes to helping a candidate. Women were more likely than men to be motivated by issue politics and to hurt the chances of a candidate, but were no different from men in wanting to help a candidate. Older respondents were more likely than younger respondents to be motivated by issues, but age was not associated with differences in motivations about affecting candidates’ chances. Ideology, Social Networks at Protest, Live in State of Convention, Race/Ethnicity, Education, and Income were not significantly related to the dependent variables in any of the equations. Given that the largest group of issue-focused protesters were concerned with issues of war and peace, I considered what would happen if this group was removed from the analysis. The results of this alternative specification are presented in Table 2. The pattern of direction and signifi- cance (or lack thereof) matches the main analysis for five of six coefficients that were the focus of the hypothesis tests. From these results, we can conclude that if issues of war and peace moved off the agenda (which, for example, was largely the case during the 2016 convention protests), the effects of partisan and independent identities would likely remain very similar. Conclusion The results of this study exhibit consistencies and incon- sistencies with the extant literature. Consistent with Klar and Krupnikov (2016), the results show that independents are much less likely than partisans to work to help a can- didate. Likewise, consistent with Heaney and Rojas (2015), the results demonstrate a trade-off between issues and party activism, as protesting at one’s own party con- vention is associated with candidate, rather than issue, 6 Research and Politics motivations. However, in contrast to the argument of Klar and Krupnikov, the results reveal that independents are more prone than some partisans to protest with the inten- tion of stopping a candidate. One possible explanation, suggested by prospect theory (Kahneman and Tversky, 1979), is that independents are more likely to join the par- tisan fray if they see the need to stop a loss (i.e. the elec- tion of a bad candidate), than to achieve a gain (i.e. the election of their preferred candidate). Thus, it may be pre- sumptuous to count independents completely out of parti- san politics. In contrast to Heaney and Rojas, the results indicate there is not necessarily a trade-off between party and issue activism when partisans are protesting their most-distant party’s convention, where partisans are often fighting for issues right beside independents. It is a mistake to see independents as having abandoned electoral politics; though, under certain conditions, they do exhibit less concern than do partisans. Yet it must also be con- sidered that a growing share of independents in the population represents an opportunity for movements. If movements can mobilize people based on issue identities – rather than partisan identities – then they may better withstand the vicissitudes of partisan politics. Advocates truly committed to issues may be more likely to push back against parties, even when their more-preferred party is in office, thus sowing the seeds for increased relevance by movements in American politics. Table 1. Probit models of protest motivation. Equation (1) Equation (2) Equation (3) Descriptive Imputation Help a candidate Hurt the chances of a candidate Express views on issue Statistics Percent of cases Coefficient (SE) Mean (SD) More-proximate party convention (= 1) 0.726*** (0.134) −0.473** (0.182) −0.365*** (0.113) 0.193 (0.011) 4.99% More-distant party convention (= 1) 0.244* (0.119) 0.564*** (0.108) 0.016 (0.087) 0.482 (0.012) 4.99% Ideology (right of conservative = 9, left of liberal = 1) 0.009 (0.026) 0.036 (0.026) −0.035 (0.021) 2.804 (0.066) 16.9% Enthusiasm about politics (almost always = 5, never = 1) 0.151*** (0.043) 0.016 (0.043) 0.052 (0.034) 3.235 (0.037) 8.49% Anxiety about politics (almost always = 5, never = 1) 0.104* (0.044) 0.187*** (0.046) 0.082* (0.035) 3.692 (0.036) 8.20% Social networks at protest (strong = 3, weak = 1) 0.004 (0.061) −0.039 (0.059) 0.087 (0.047) 1.895 (0.026) 4.78% Live in state of convention (= 1) 0.013 (0.119) 0.206 (0.118) −0.119 (0.090) 0.779 (0.014) 13.34% Sex/Gender (female = 1) 0.092 (0.097) 0.190* (0.093) 0.301*** (0.075) 0.495 (0.017) 7.92% Race/Ethnicity (white = 1) −0.112 (0.135) −0.010 (0.142) −0.106 (0.104) 0.884 (0.011) 6.49% Age (in years) −0.001 (0.004) −0.002 (0.003) 0.013*** (0.003) 38.159 (0.525) 8.56% Education (graduate degree = 6, less than high school = 1) 0.025 (0.037) 0.000 (0.037) 0.050 (0.030) 3.992 (0.045) 8.63% Income (thousands of dollars) −0.001 (0.001) −0.001 (0.001) −0.001 (0.001) 50.118 (1.821) 11.98% Constant −2.370*** (0.334) −2.158*** (0.327) −1.330*** (0.245) N 1377 1377 1377 Mean of the dependent variable 0.145 0.074 0.316 Strata 2 2 2 F(12, 1364) 5.130*** 7.650*** 8.170*** SD: standard deviation; SE: standard error. *p ⩽ 0.05. **p ⩽ 0.01. ***p ⩽ 0.001. Heaney 7 Looking to the future, both the Democratic and Republican parties in the United States are facing increased contention from social movements – such as the Tea Party, Occupy Wall Street, and Black Lives Matter – some of which is likely to manifest itself through convention pro- tests. Politically oriented scholars would be well-advised to be attentive to these dynamics in order to understand how organized (and disorganized!) groups may be disruptive to party coalitions. Acknowledgements Data for this paper were collected as a joint effort with Dara Strolovitch, Joanne Miller, and Seth Masket. The author appreci- ates the generosity of his collaborators in allowing the use of the data for this project. An earlier version of this article was presented at the William and Flora Hewlett Foundation Conference on “Parties, Polarization and Policy Demanders”, University of Maryland, College Park, 9–10 June 2016. For helpful comments, thanks are owed to Stephanie Evans, Stephen Garcia, Ethan Kaplan, Ken Kollman, Michael Martinez, Lilliana Mason, Rob Mickey, Nurlan Orujlu, Zoe Van Dyke, and Winnie Wang. Declaration of conflicting interests None declared. Funding This work was supported by the National Science Foundation, Small Grants for Exploratory Research (grant numbers 0842474 and 0842371) and the Center for Urban and Regional Affairs at the University of Minnesota. Table 2. Probit models of protest motivation – no antiwar protesters. Equation (1) Equation (2) Equation (3) Descriptive Imputation Help a candidate Hurt the chances of a candidate Express views on issue Statistics Percent of cases Coefficient (SE) Mean (SD) More-proximate party convention (= 1) 0.803*** (0.152) −0.421* (0.209) −0.327* (0.154) 0.213 (0.012) 5.50% More-distant party convention (= 1) 0.209 (0.141) 0.663*** (0.128) −0.122 (0.119) 0.380 (0.014) 5.50% Ideology (right of conservative = 9, left of liberal = 1) −0.005 (0.030) 0.029 (0.029) 0.073** (0.026) 3.070 (0.063) 18.57% Enthusiasm about politics (almost always = 5, never = 1) 0.145** (0.047) −0.015 (0.049) 0.027 (0.045) 3.080 (0.037) 8.74% Anxiety about politics (almost always = 5, never = 1) 0.073 (0.050) 0.181*** (0.052) 0.108* (0.047) 3.589 (0.035) 8.94% Social networks at protest (strong = 3, weak = 1) 0.023 (0.070) 0.024 (0.070) 0.169** (0.065) 1.915 (0.025) 5.40% Live in state of convention (= 1) −0.025 (0.135) 0.122 (0.135) −0.475*** (0.112) 0.721 (0.013) 14.73% Sex/Gender (female = 1) 0.144 (0.113) 0.183 (0.112) 0.320** (0.104) 0.435 (0.015) 8.64% Race/Ethnicity (white = 1) −0.097 (0.153) −0.067 (0.161) −0.243 (0.132) 0.845 (0.011) 7.17% Age (in years) 0.001 (0.004) −0.002 (0.004) 0.008* (0.003) 35.661 (0.469) 9.43% Education (graduate degree = 6, less than high school = 1) 0.028 (0.043) 0.012 (0.045) 0.032 (0.039) 3.820 (0.042) 9.43% Income (thousands of dollars) −0.001 (0.001) −0.001 (0.001) −0.003* (0.001) 46.455 (1.730) 12.87% Constant −2.316*** (0.369) −2.181*** (0.378) −1.786*** (0.310) N 998 998 998 Mean of the dependent variable 0.119 0.137 0.173 Strata 2 2 2 F(12, 985) 4.42*** 5.94*** 4.79*** SD: standard deviation; SE: standard error. *p ⩽ 0.05. **p ⩽ 0.01. ***p ⩽ 0.001. 8 Research and Politics Supplementary material The replication files are available at: https://dataverse.harvard. edu/dataverse/researchandpolitics. The supplementary files are available at: http://rap.sagepub. com/content/3/4 Notes 1. The survey questions in this analysis are listed in Supplement 1. This protocol received institutional review board (IRB) approval and all participants gave informed consent. 2. A summary of responses to all motivations is provided in Supplement 2. 3. Respondents were counted as having issue motivations if they wrote a policy issue in the space provided. If they wrote something other than a policy issue (e.g. “to meet people”), then they were counted as a zero on this variable. 4. Of those who said that they sought to hurt a candidate, 29.6 percent sought to help a candidate and 48 percent sought to promote an issue. Of those who sought to help a candidate, 37.6 percent sought to hurt a candidate and 35.4 percent sought to promote an issue. Of those who sought to promote an issue, 10.3 percent sought to help a candidate and 17.8 percent sought to hurt a candidate. Carnegie Corporation of New York Grant The open access article processing charge (APC) for this article was waived due to a grant awarded to Research & Politics from Carnegie Corporation of New York under its “Bridging the Gap” initiative. The statements made and views expressed are solely the responsibility of the author. References Bawn K, Cohen M, Karol D, et al. (2012) A theory of political parties. Perspectives on Politics 10(3): 571–597. Clemens E (1997) The People’s Lobby. Chicago, IL: University of Chicago Press. Corrigall-Brown C (2012) Patterns of Protest. Stanford, CA: Stanford University Press. Edelman M (1964) The Symbolic Uses of Politics. Urbana, IL: University of Illinois Press. Frymer P (1999) Uneasy Alliances. Princeton, NJ: Princeton University Press. Heaney MT and Rojas F (2007) Partisans, nonpartisans, and the antiwar movement in the United States. American Politics Research 35(4): 431–464. Heaney MT and Rojas F (2014) Hybrid Activism. American Journal of Sociology 119(4): 1047–1103. Heaney MT and Rojas F (2015) Party in the Street: The Antiwar Movement and the Democratic Party after 9/11. New York: Cambridge University Press. Hermes K (2015) Crashing the Party. Oakland, CA: PM Press. Kahneman D and Tversky A (1979) Prospect theory. Econometrica 47(2): 263–292. King G, Honaker J, Joseph A, et al. (2001) Analyzing incomplete political science data. American Political Science Review 95(1): 49–69. Klar S and Krupnikov Y (2016) Independent Politics. New York: Cambridge University Press. Layman G, Carsey T, Green J, et al. (2010) Activists and con- flict extension in American party politics. American Political Science Review 104(2): 324–346. Little R (1988) Missing data adjustments in large surveys. Journal of Business and Economic Statistics 6(3): 287–296. Magleby D and Nelson C (2012) Independent leaners as policy partisans. The Forum 10(3): Paper 6. Polletta F and Jasper J (2001) Collective identity and social move- ments. Annual Review of Sociology 27: 283–305. Schlozman KL, Verba S and Brady HE (2012) The Unheavenly Chorus. Princeton: Princeton University Press. Schattschneider E (1960) The Semisovereign People. New York: Holt, Rinehart & Winston. Schlozman D (2015) When Movements Anchor Parties. Princeton, NJ: Princeton University Press. Skocpol T and Williamson V (2012) The Tea Party and the Remaking of Republican Conservatism. New York: Oxford University Press. Sobieraj S (2011) Soundbitten. New York: New York University Press. Staff (1908) For anti-injunction plank. New York Times, 9 June, 2. Walgrave S and Verhulst J (2011) Selection and response bias in protest surveys. Mobilization 16(2): 203–222. https://dataverse.harvard.edu/dataverse/researchandpolitics https://dataverse.harvard.edu/dataverse/researchandpolitics http://rap.sagepub.com/content/3/4 http://rap.sagepub.com/content/3/4