From the beginning, political science has focused on the study of political behavior. Citizens were deemed to be voters who could be analyzed and understood with the proper data. The study of voters centered on single-country investigations (Campbell et al. 1960) but quickly moved toward comparative analyses of multiple nations (Almond and Verba Reference Almond and Verba1963) while also underlining the relevance of different types of voters and political behaviors. Once scholars expanded their research to newer forms of political action, the realm of conventional activism appeared limited. Newly inspired by the surrounding political context, political scientists chose to investigate political-protest activities as legitimate forms of political behavior. Gurr (Reference Gurr1968) and Marsh (Reference Marsh1977) were among the first examples of single-country studies on protest, followed soon after by limited comparative research on political contention (Barnes and Kaase Reference Barnes and Kaase1979). More recently, the relevance of protest activism in politics has led to many comparative investigations with a large-N approach for a clear effort toward a systematic comparison of protest action across countries and traditions (Dalton, van Sickle, and Weldon Reference Dalton, Sickle and Weldon2010; Inglehart and Catterberg Reference Inglehart and Catterberg2002; Jennings and van Deth 1989).
As scholars worked to better understand who protested and how and why individuals embraced this type of political engagement, it became clear that the lack of data was the most serious obstacle, especially in cases of comparative research. For the first few decades, scholars of protest had to collect their own data because interest in unconventional political behavior was not shared by most researchers. For a long time, understanding voters was more important in political science because it was easier to identify individuals who voted versus citizens who participated in a protest activity. The study of voters was more relevant, better funded, and more engaging for political scientists. Investigations concerning protesters were deemed generally less trendy, less financially supported, and certainly less mainstream. In some cases, research on unconventional mobilization in politics was viewed as a project designed more by sociologists than political scientists.
In the past three decades, the typical concern with data availability in protest studies has been addressed successfully. Multiple data-collection programs acknowledged the relevance of unconventional political behavior in research concerning political engagement, social movements, and citizen activism (Dalton Reference Dalton2014; Della Porta Reference Della Porta2015; Rucht Reference Rucht, Dalton and Klingemann2007; Schussman and Soule Reference Schussman and Soule2005). However, as one concern was dismissed, two new issues came to the forefront. Researchers involved in the study of protest still had to address two challenges: recurring problems with the understanding of the concept “protest” and its most appropriate “measure.” In particular, confusion and vagueness related to the concept and measure of protest became evident when comparative studies tried to predict political confrontation across countries with dissimilar interpretations of protest activism.
Even after decades of research, scholars of protest fail to agree on the actual definition of the concept Footnote 1 that they want to measure and its best possible empirical quantification. Best practices in research highlight the need to clearly identify the concept that is being investigated as a first step toward the selection of its best possible measure.
A FLUCTUATING CONCEPT
The literature on protest consistently reveals the struggle with identification of the concept, beginning with the actual terms used to single out this type of political action. Initial examples demonstrated the general confusion about what political protest was: civil strife (Gurr Reference Gurr1968), unorthodox political behavior (Marsh Reference Marsh1977), and political violence or simply unconventional political activism (Barnes and Kaase Reference Barnes and Kaase1979). Subsequent research added many more terms to categorize the actions belonging to the newest repertory of political activism. Among the more common terms, scholars used peaceful confrontation, nonviolent activism, unconventional engagement, and nonconfrontational political behavior. This terminology highlights not only the diverse understanding of the concept but also the corresponding empirical problem in the research to convincingly measure an ever-changing concept. For example, in the beginning, political violence was easily included as an example of protest: riots, property destruction, and personal violence for political purposes were all considered episodes of protest activity (Barnes and Kaase Reference Barnes and Kaase1979). However, as scholars focused on measuring the events they described to be able to collect the data they needed, it became clear that the identification of activists was too difficult and sometimes dangerous. When the research included more examples of extreme confrontation as part of the concept of protest, data availability suffered. As a consequence, subsequent studies quickly simplified the meaning of protest, concentrating on unconventional activism that was less dangerous, more feasible, and potentially more likely to be embraced by citizens. Current large surveys Footnote 2 on political behavior (e.g., the World Values Survey and the International Social Survey) include more practical and less intimidating examples of protest: signing a petition, participating in a boycott, and taking part in a peaceful demonstration or a strike. At times, these same surveys incorporate the “other protest” category to include possible events of more challenging (and perhaps illegal) confrontational activism.
Among the more common terms, scholars used peaceful confrontation, nonviolent activism, unconventional engagement, and nonconfrontational political behavior. This terminology highlights not only the diverse understanding of the concept but also the corresponding empirical problem in the research to convincingly measure an ever-changing concept.
The changes in the possible list of acts of protest affected the interpretation of the concept. For some researchers, protest can only be violent to be unconventional. For other scholars, anything that is considered confrontational (whether or not violent) belongs to the repertory of unconventional activism—or, stated more simply, any political act other than voting, writing to a representative, or volunteering for a political campaign. This lack of specificity in the concept definition remains a constant feature in the research on protest, and it is equally reflected in the three contributions to this symposium. The evolution in the understanding of the concept of protest affected the corresponding meaning of protest activities. Ultimately, this meant problems for comparative studies when authors presented a long-term assessment of how citizens embraced more unconventional activism in their repertory of political-behavior actions.
The clear conclusion from all of the research so far is that protest is a dynamic concept (Hosch-Dayican Reference Hosch-Dayican2014). The meaning of protest depends first on the country where it is studied because street demonstrations, for example, are simply more frequent in France than in England. Footnote 3 Second, unconventional activism is linked to the period considered because occupying a building was more in vogue in the 1970s, whereas signing a petition currently is trendier with online social-media activism. Footnote 4 Third, confrontational engagement relies heavily on the type of technology available. Random tire-deflating campaigns required only a sharp tool but also were time consuming, whereas Twitter attacks are more feasible, effective, and able to focus on specific and relevant targets Footnote 5 rather than random car owners. Although scholars of political activism are engaged in creating updated conceptual maps of newer forms of political behavior (van Deth Reference van Deth2014), the conceptual problem surrounding protest has remained, adding a findings-comparability issue across publications.
The most recent addendum to the ongoing conceptual confusion around protest developed progressively in the past decade. If disciplines such as political science and sociology were generally at the forefront of the study of confrontational activism, the latest innovation in communication technology has spiked investigations on protest participation from media and communication-studies scholars (Shirky Reference Shirky2011; Valenzuela Reference Valenzuela2013). Recent approaches to the study of protest have added new understanding to the concept of unconventional behavior, including an innovative subfield of research on “digital activism.” At this time, protest is not only a physical manifestation of political behavior (i.e., street demonstrations, building occupations, and strikes); it is equally an online political activity in which political boycotts against candidates, campaigns, parties, and companies are used as forms of political activism. Even more important is the fact that this new realm of protest (i.e., online activism) is a separate dimension of political behavior (Gibson and Cantijoch Reference Gibson and Cantijoch2013).
The ultimate example of how protest has become an online activity—that is, successfully selected as a possible act of political engagement—is the focus in democratization studies on citizens’ social-media use as a form of protest activism for cases of democratic revolutions (Diamond and Plattner Reference Diamond and Plattner2012). If the old concept of protest required individuals to be recruited in protest activities based on their own personal and physical involvement, the current emphasis in identifying and collecting data on protest cases is on a democratic movement’s selection and use of Facebook, Twitter, or YouTube to carry out unconventional political attacks on totalitarian regimes. The advantage of this additional interpretation of protest is that it is easier to record and gather data because any individual’s digital footprint can be shared quickly (although its interpretation is not as simple).
The concept of protest started as an unacceptable version of political participation, used only by citizens who wanted to reject their own democratic government (e.g., the antisystemic activists in the 1960s). As the idea of protest evolved, scholars attempted to study it more empirically. The new unconventional-activism concept developed into a version of refined and sophisticated political engagement, in which citizens with a strong political interest would embrace more challenging and risky political activities (e.g., street demonstrations and building occupations) as examples of their strong commitment to political participation outside of the regulated realm of voting. Recently, the concept of protest has been substantially revamped to fit current times: protesters today are political activists involved in sharing information about political boycotts or online petitions through different social-media platforms. In some cases, these new online activists also embrace a riskier version of online unconventionality by becoming hackers for specific political reasons (Beyer Reference Beyer2014). After more than a half-century of studying unconventionality, the concept remains multifaceted, with more variation in interpretations of the term “protest” in the scholarship.
With new authors studying different forms of political behavior across countries, from voting to political violence, Teune (1968, 126) asked the important methodological question that still unsettles comparative-politics scholars: Can we really compare? If so, which countries could researchers of protest behavior include in the comparison?
AN INCOMPARABLE MEASURE
For scholars of protest behavior, an empirical understanding of a fluid concept quickly became equally problematic. Teune (Reference Teune1968) acknowledged “the problem of measurement” when studies of comparative political behavior began to appear in the discipline. With new authors studying different forms of political behavior across countries, from voting to political violence, Teune (1968, 126) asked the important methodological question that still unsettles comparative-politics scholars: Can we really compare? If so, which countries could researchers of protest behavior include in the comparison? This type of research dilemma potentially can be solved for small-N studies, when scholars could opt to collect their own data. Yet, the same issue of validity of the actual measure remained clear when investigators had to compare the same indicators across many countries, groups, political systems, and levels of analysis. Any measure used to represent protest likely would lose its validity because citizens from different nations embrace a diverse set of examples of protest and to a different extent.
Similarly, the issue of comparability has always been at the forefront of the comparative-politics subfield but, for protest studies, the problem surrounding the actual empirical measure was even more concerning than the same issue regarding its concept. Naturally, the first step would be to select the same indicator of protest across the sample of units in the study. For instance, a street demonstration could be considered an ideal setting to measure how citizens embrace unconventionality in politics. Even with such well-known events, governmental rules about demonstrations vary across political and legal systems. In countries such as France, demonstrations are not only easily accepted, used, and expected, they also are practical and fairly popular across all societal groups. In other countries, demonstrations instead are more difficult to organize due to legal requirements and stricter police supervision. If street marches happen in countries that are only partially democratic, they tend to be less popular and more sporadic, immediately creating a sample-quality issue for investigators gathering data.
A second step in the study of confrontational activism is related to the selection of a group of equivalent items. If identical indicators of protest across groups of cases do not exist or cannot be reliably collected, then a better measure of the dimension of protest was a group of empirical items relating to one another. In this situation, studies of protest across countries included a set of protest actions that effectively measured unconventionality, were fairly feasible across different cases, and also were related. The 1970s Political Action study (Barnes and Kaase Reference Barnes and Kaase1979) was the first research project to assess competitive items as measures of protest. Footnote 6 Later, large-survey programs followed this possible solution about the incomparability of protest and began asking questions about petitions, peaceful demonstrations, boycotts, and strikes. This decision provided more data availability for scholars of protest and spiked the number of studies in which multiple items of unconventionality are scaled together into a single index of protest. Przeworski and Teune (Reference Przeworski and Teune1967) suggested this approach when discussing issues of validity for protest measures. With the multiplication and advancement of statistical software, scales of protest gained in popularity among scholars. Footnote 7 However, this approach led to the need for possible competing scales of protest across countries because the reliability of the index may vary among cases. Footnote 8
A third step to solve the dilemma about a valid and comparable protest measure is linked to the new interpretation of protest that has recently spread through communication-studies research on political activism. Examples of this new repertory of protest measures include datasets on protest events worldwide and access to data on the number of tweets per hour or day. The first mode of new protest is represented by the open-access GDELT Footnote 9 project, organized and supported by Google. The new technology company has set up a website where researchers can look up and download data on the actual number of global protest events, catalogued using online media sources of different types. The appeal of this type of empirical measure of unconventionality is clear. First, the actual database is extremely large, occasionally causing problems in data handling. Yet, because of the vast amount of raw data, researchers can afford to be selective relative to which countries to compare because there are many competing possibilities and a significant amount of data collected. This situation provides multiple possible protest measures that could work across a different set of countries. Second, this online archive is updated frequently, allowing scholars to investigate the most recent events, without having to wait—generally at least a year—for the official data release. Third, it is free, which helps new researchers and institutions without much funding by giving them the ability to contribute to the empirical analysis of protest behavior at a time when street marches have become frequent challenges to government policies. In brief, if it is not already apparent, there is an urgent need for scholars across disciplines to study protest.
The second mode of new protest is provided by another new technology company: Twitter. In this case, scholars can buy data on the number of tweets related to a specific hashtag during a certain period. Investigators can study a trend as it evolves online (e.g., number of tweets per hour) and can access the actual time development of a digital protest movement (Freelon, McIlwain, and Clark Reference Freelon, McIlwain and Clark2016). The company has the ultimate decision-making power in granting availability of the data, and it does not always grant all of the requests submitted. In this context, where protest meets technology, scholars of new and innovative forms of protest activism are simply trying to catch up. The new repertory of protest deals with large-scale online calls to marches for a political cause as well as identification of potential supporters as online activists (from bloggers to hackers) to oppose the government.
The second mode of new protest is provided by another new technology company: Twitter. In this case, scholars can buy data on the number of tweets related to a specific hashtag during a certain period. Investigators can study a trend as it evolves online (e.g., number of tweets per hour) and can access the actual time development of a digital protest movement (Freelon, McIlwain, and Clark Reference Freelon, McIlwain and Clark2016).
Overall, scholars across multiple disciplines now have access to old and new measures of protest, which has increased the quantity and quality of the scholarship on unconventional political behavior. Concerning issues of validity and comparability, this development has not settled the question about which measure in the research on protest is dominant. More measures have brought more possibilities but still no consistency.
A BRIEF OVERVIEW OF THE SYMPOSIUM CONTRIBUTIONS
In light of the many facets of protest, as a concept and a measure, the three contributions to this symposium use different examples of protest to represent the various actions linked to political unconventionality. In particular, the articles highlight how protest can be considered essentially a one-item measure only (i.e., demonstrations for Matt Schoene); a three-separate-items dimension (i.e., petitions, demonstrations, and boycotts for Mario Quaranta); and two competing scales (i.e., low-cost versus high-cost indexes for Marc Guinjoan and Toni Rodon). This type of approach confirms the ongoing dilemma related to the measure of protest. If a single-item measure is considered easier to collect and simpler to interpret, then a multiple-item scale adds complexity to the understanding of a multifaceted concept and a more thorough representation of the hybrid activism currently associated with protest. Yet, as always, the more items in a scale, the less likely it is for that scale to be useful in a cross-country comparative analysis.
On one hand, the authors in this symposium are responding to the need to provide a more comprehensive representation of what protest is today. On the other hand, their contributions also are underlining how a more moderate size for the sample of countries in the study is needed when large-N comparisons on the topic of protest are investigated. All of the articles are using several waves of the European Social Survey program as the source of protest actions included. In their research to test hypotheses linked to confrontational political action, all of the authors were inspired by the protest movements and large-scale political action against governments in Europe after the 2008 global financial crisis.
Schoene’s contribution on urban protest focuses on the role of cities (i.e., large urban areas) as incubators for protesters and confrontational activism, regardless of the specific focus of the protest-action movement investigated. His article argues for a need to focus on the role of cities as equally important predictors of protest as the more traditional variables (i.e., grievances, resources, and causes) usually present in the more qualitative literature on urban protest movements. This type of conclusion is based not only on the current expansion of protest activities in cities or urban population growth. Cities also matter in protest studies because of their advantage relative to organizational spaces, meeting opportunities, and synergy from different forms of activism.
In considering the relevance of how protest can be organized, Quaranta’s article centers on the role of more traditional political organizations (i.e., political parties and trade unions) in instigating and supporting unconventional political activity. Recent research on the role of more traditional political groups painted a pessimistic picture about the less-relevant contributions of parties and trade unions toward more engaging forms of activism. If the literature on protest movements highlights the innovative contribution from new and much less rigidly structured movements, Quaranta’s analysis presents evidence about the still-important meaningful task performed by parties and trade unions regarding demonstrations, boycotts, and petitions when the economy is doing poorly. As predicted by deprivation theory, a downturn in the economy encourages unconventional political activism; however, citizens who are party and union members are more likely to choose different forms of protest than individuals who are not part of those traditional political organizations. In some ways, this is revenge for the old traditional groups in politics, in light of the much stronger evidence for new types of more fluid movements, without a clear membership.
The third article by Guinjoan and Rodon discusses the possibility of a protest gap that finally has been closed after the more recent increases in protest-activity cases. The authors consider measures of both high-cost and low-cost protest in their assessment of whether individuals are equally likely to engage in confrontational activism after 2008. The protest gap long debated in the literature seems to have been closed only in cases of low-cost unconventional activism (i.e., petitions and boycotts). Yet, more-demanding opportunities to become politically active reveal an even larger protest gap, with individuals from different groups having different probabilities of participating in street demonstrations. A gap in unconventional activism persists across nations, even after the most serious global financial crisis provided cause for many individuals to more loudly express themselves politically.
These three contributions to this symposium represent three views of protest. The articles highlight how the actual concept and measure of political unconventionality are as equally diverse as they always have been and how advances in empirical methodology have supported scholars in their ongoing study of this mode of political activism. Protest remains a highly important example of political behavior, and its spread across countries in the past half-century demonstrates the need to keep focusing on political boycotts as much as voting. In particular, since the 2008 global economic collapse, scholars of protest have addressed new forms and modes of political engagement that would not have been studied otherwise. In this context, our hope is that this symposium will begin a conversation about exploring protest more consistently and with regard to multiple new measures to come.
ACKNOWLEDGMENTS
As the special guest editor for this symposium focusing on protest, I thank the four authors for their contributions; the editors of PS: Political Science & Politics; and especially Phillip Ardoin and Celina Szymanski for their support, patience, and encouragement. Some of the content in this symposium was presented in part at either the 2016 European Social Survey Conference in Lausanne or the 2016 APSA Annual Meeting in Philadelphia. Special recognition goes to Professor Samuel H. Barnes, a teacher and scholar of political protest.
