Abstract
This paper examines how the use of social media affects participation in offline demonstrations. Using individual web survey data from Norway, we ask whether social media usage serves to re-affirm or transcend socioeconomic divides in participation. In addition to data on demonstration participation in general, we also use the data on the Rose Marches that were organized after the 22/7 terror events as a critical case. Our results show that the type of participant mobilized via the social media is characterized by lower socioeconomic status and younger age than those mobilized via other channels. We also show that connections to information structures through social media exert a strong and independent effect on mobilization. Our findings thus appear to corroborate the mobilization thesis: social media represent an alternative structure alongside mainstream media and well-established political organizations and civil society that recruit in different ways and reach different segments of the population.
Introduction
In this paper we examine the impact of social media usage on participation in civic and political offline demonstrations and discuss whether they work to consolidate or transcend social and cultural divides. An increasing amount of research has been concerned with the rise of new inequalities resulting from the growth of the internet under the label of ‘digital divide’ (Norris, 2001; Van Dijk, 2005). As access to the internet has increased in many countries, efforts have been made to re-conceptualize digital inequality into three discrete levels: access to technologies; digital proficiency; and propensity to take advantage of technological affordances (Stern et al., 2009). The last type of digital inequality concerns, among other things, whether people make use of the internet and social media for political and civic ends. Discussing the state of research in this field, DiMaggio et al. (2004) concluded that users from privileged backgrounds were more likely to use the internet for community purposes than people at the lower end of education and income scales (2004: 379). Recent studies also indicate that online proficiency and use co-vary with background factors, such as geographical location (Stern and Adams, 2010) and socioeconomic status, and that this holds even among young people (Hargittai, 2010).
This paper analyses the impact of social media on a particular form of civic activity, the offline demonstration. When it comes to the relationship between online and offline participation, research has produced diverging results. Some researchers (Bimber, 2001; Brants et al., 1996; Streck, 1997) conclude that online participation has no or only a very limited effect on offline activity, whereas others (Bakker and De Vreese, 2011; Jennings and Zeitner, 2003; Tolbert and Mcneal, 2003; Weber et al., 2003) find a significant relationship between online and offline forms of political participation. However, studies of the role of the internet on civic engagement have so far tended to look at the impact of traditional internet websites (so-called Web 1.0) on political engagement. In this paper we argue that social media platforms (Web 2.0), such as Facebook, Twitter and YouTube, have particular characteristics – or affordances – that distinguish them from Web 1.0. and which may alter the impact of social media on offline civic and political activity.
By ‘affordance’ we refer to the type of action or a characteristic of actions that a technology enables through its design. Typical for the affordances of social media is that they allow for the establishment of a social network structure and for interactive exchange within and between such defined networks (boyd, 2011). In line with Norris (2002), we conceive of mobilization processes as dependent both on individual characteristics, such as economic resources, motivation, and skills (individual agency), and on the structures facilitating the spread of information and motivation of individuals to participate (mobilizing agency). Social media affordances may change both the individual and the mobilizing agency aspects of mobilization processes by introducing new types of communication structure and by allowing for new communication forms. The central element of this transformation is the network structure that underpins social media. Consistent with the ‘strength of weak ties’ argument (Granovetter, 1973), social networks have been shown to play a crucial role in politics inasmuch as they shape the way political information is received and interpreted (Huckfeldt and Sprague, 1987, 1995; McClurg, 2006). Similarly, larger online networks have been shown to be associated with online civic participation (De Zúñiga and Valenzuela, 2011; Valanzuela et al., 2011).
In what follows we first describe the research context and present the case of what became known in Norway as the Rose Marches. We then present a perspective on how social media may transform individual and mobilizing agency in processes of mobilization. The empirical part of the paper asks first whether groups mobilized through social media differ from those mobilized through conventional channels. We go on to analyze the impact of social media on the individual and mobilizing agency dimensions of mobilization by looking at informational and motivational roles of social media in demonstrations and by analyzing the relative roles of social media and offline political and civil society organizations as structures for mobilization. The analysis of these two dimensions of mobilization puts us in a better position to understand the transformations entailed by social media in terms of transcending participatory divides.
Research context
We use Norway as our case to discuss the impact of the use of social media on participation in political and civic demonstrations. In a comparative perspective Norway is characterized by high access to technology. In 2009, 99% of families with children and 82% of families without children possessed a PC (personal computer) (SSB, 2011). Ninety-two per cent of the population was connected to the internet at home, reaching almost 100% among people under 45. Given the absence of a digital divide in terms of internet access, Norway is a particularly apt case for studying the capacity of social media to transcend the social and cultural divides in political participation. Is there a class-related difference in the propensity to take advantage of the technology for civic purposes, given the ubiquity of technological access, or not?
We use individual web survey data to assess whether social media usages are changing the manner in which people are mobilized to take part in offline demonstrations. The data allow us to examine the propensity of individual citizens to take part in demonstrations in general, and in the Rose Marches in particular, held in Norway in the aftermath of the 22/7 terror attacks. The Rose Marches sprang from an idea posted on Facebook to commemorate the victims of the killings at Utøya and express people’s revulsion at the act. Within three days of the attack marches were being held in cities, towns and communities throughout the country. As many as 34% of our respondents had participated in one of these marches.
We use the Rose Marches as an illustration of a broad mobilization process in Norway. While the dramatic background makes it rather special as a sociological case, at the same time it displays ideal typical elements of mobilization processes facilitated by the social media. The Rose Marches were also held in defense of certain civic values (Wollebæk et al., 2012). After starting as a ‘Facebook event’, word spread quickly across the medium and was eventually backed by well-established civil society organizations such as Amnesty International Norway (Wollebæk et al., 2011). The Rose Marches may therefore function as a natural experiment, allowing us to look in more detail at the relative roles of different media and of established civil society structures.
The properties of social media as channels for mobilization: Affordances and networks
Social media sites, we argue in this paper, have distinct, inherent properties, conceptualized as affordances and network functionalities. These properties reduce the cost of civic and political participation very considerably. Indeed, the resources required by political participation are usually expressed in terms of time, money, and civic skills, such as communication and organizational capacities (Brady et al., 1995). As online technology develops, and the cost of information retrieval and communication in general falls, political participation is expected to become less costly (Benkler, 2006; Sylvester and McGlynn, 2010). As a result, patterns of mobilization could be transformed, both in terms of who participates and how they participate.
However, as pointed out by DiMaggio et al. in relation to Web 1.0, this point of view is open to question since many information sources were already free or inexpensive, and since differences in skills and civic engagement may well persist despite the social media (2004: 371). In order to determine whether social media do in fact narrow or even render these divides redundant, we need to take the specific affordances of social media into account, that is their profiles, friend lists and tools of communication (boyd, 2011). Based on these affordances social media have a particular capacity to link people within a digital network. One of the powerful achievements of such networks is in bridging the local and the global, allowing local phenomena to be spread across the entire network and produce global effects. Digital networks enable social processes where individual behaviors are aggregated to produce collective outcomes. Information cascades are one of those social processes that occur when people make decisions sequentially, are able to observe others’ decisions, draw rational inferences from these decisions, and imitate them on the basis of their inferences. The small-world network structure of social media (Watts, 1999) is conducive to information cascades because users can easily observe what their connections do, make inferences and decisions on the basis of these observations, which in turn are propagated further along the network.
The networking functionalities and affordances of social media make them a potentially very efficient means of spreading information among a population about forthcoming actions and demonstrations. Not only do information cascades have an informational effect, they arguably have a motivational effect as well in that social media publicize people’s decisions to join a group or sign up for an event. Political scientists and communication scholars have demonstrated how the use of the new media facilitates a more positive and active outcome of political behaviors and participation. However, the nature of the correlation between media use and political behavior is still a contentious issue (see Kim and Kim (2012) for a review). Social media, in conveying more ‘personalized’ information arising from strong and weak ties within the individual’s social network, are likely to enhance the relevance of political information as well as motivating people to act on the basis of this information.
The impact of social media on mobilization processes: Individual and mobilizing agency
We follow Norris (2002) in perceiving civic and political participation as the joint result of individual and mobilizing agencies. The most cited model of individual agency as a basis for mobilization is Verba et al.’s (1995) civic voluntarism model. It explains individual differences in civic and political participation with respect to individual inequalities in resources (income, education, skills) and motivation (political interest, information, self-efficacy) that characterize citizens’ social and political background. Theories of mobilizing agency (Rosenstone and Hansen, 2003), on the other hand, seek to explain the role of organizations and social networks, such as political parties, trade unions, voluntary associations, and informal networks, in mobilizing, engaging and organizing citizens. Mobilization is defined as ‘the process by which candidates, parties, activists and groups induce other people to participate’ (Rosenstone and Hansen, 2003: 25). Social networks on this account are facilitators of political mobilization. Through ties of camaraderie, neighborliness, and family, social leaders are able to communicate their messages and engage people in civic and political action.
With the emergence of social media, the conceptions of individual agency and of mobilizing agency as a basis for mobilization are in need of modification. We would expect social media to make an impact on both counts. On the individual side, we need to consider two dimensions – motivation and resources. In particular, social media arguably reduce the impact of resource inequalities on civic and political engagement, since social media offer new and open types of information networks alongside the networks embedded in organized civil society. Research on political participation has indeed indicated a link between high socioeconomic status and protest participation (Barnes and Kaase, 1979). Recent multilevel analyses of European social survey (ESS) data confirm that unconventional participation is more contingent on high education and income than voting (Berglund et al., 2008; Schäfer, 2010). In the specific case of demonstrations, however, more recent studies suggest instead declining social differences. Norris et al. (2005) argues that as protesting has become a legitimized tool of political action, the demonstrators’ socioeconomic characteristics are less distinct than before. On the other hand, Gallego’s (2007) analysis of ESS data shows significant effects of education in most countries, while income and class yield contradictory and mostly insignificant results. Studies of the Norwegian case seem to conform to this pattern (Gallego, 2007; Olsen and Sætren, 1980; Pettersen and Rose, 1996).
With regard to mobilizing agency, we would expect digital networks to play an increasingly important role as channels of mobilization, alongside traditional mobilizing agencies. It has been argued that social media will make organizations less necessary by dramatically reducing the costs of organizing collective action (Shirky, 2008). In Norway, civil society organizations have played an important role as structures for spreading information and mobilizing participation. It is therefore interesting to ask whether the emergence of social media as a mobilizing structure has affected previously institutionalized structures, and if so, how. Social media affordances and network logic can be leveraged by both the political and civic establishments: political leaders, political parties, and voluntary organizations. On the other hand, social media affordances and network logic may also be leveraged by individual citizens wanting to encourage other citizens to take action on an issue they deem important.
Methods and data
The data on which the analyses below are based consist of a two-wave web-survey carried out in April and August 2011. The samples were drawn from TNS Gallup’s web panel, comprising 62,000 individuals who are representative of the population of internet users. The first data collection was undertaken in March/April 2011 and consisted of three parts: a sample representative of the internet population aged 16 and over (N=1127); a sample representative of the population of active social media users (N=4183); and an extra sample of individuals aged between 16 and 24 (N=427). The second wave was undertaken between August 12 and August 17, 2011 and designed to capture the role of social and other media in the weeks after the terrorist attacks, as well as possible attitudinal and value changes in the population as a result of them. The August data consisted of two sections: (1) a separate population sample (N=931); and (2) re-interviews of 2252 of active social media users (Facebook > once a week or Twitter ≥ once a week) who were interviewed during the first wave in April.
The response rates of the population samples, defined as the percentage of contacts made resulting in valid responses, were 48% at both times. Sixty-six per cent of the respondents from the first wave who were contacted again responded to the second questionnaire. Design weights were added to make the material representative of the web population.
In our analysis we used the following variables to capture the different dimensions of mobilization processes.
The core dependent variable was participation in demonstrations. In the April/May survey respondents were presented with a list of different political and civic activities and asked to indicate whether or not they had participated in any of them during the past six months. Participation in an offline demonstration was one of the items on the list. In the August survey, respondents were asked specifically whether or not they had participated in Rose Marches or related events of any kind.
In order to capture how social media facilitated people’s participation in demonstrations on the motivation dimension, we needed to assess how social media were used to learn of impending demonstrations. A question about how demonstration participants first received information about a demonstration was used to compare the informational value of different types of media and personal contact. In relation to the Rose Marches, we also developed a set of more detailed questions about the use of social media and mainstream media for accessing information about events and for more interpersonal purposes such as expressing sympathy.
When analyzing the social network dimension of mobilization, we take it as read that organized civil society constitutes an established network by which information can be dispersed among members, and that social media may constitute an alternative structure. In order to assess the impact of these structures, we assessed the embeddedness of participating individuals in these two structures by measuring the number of memberships in voluntary organizations and in Facebook groups respectively.
The resource dimension was measured using the following individual background variables: age, education level, gender, household (and, in separate analyses, personal) income, and urbanity.
Are new groups mobilized through social media?
The main question posed by this paper is whether the use of social media reduces or eliminates social and cultural divides in participation in offline demonstrations. In line with the situation in other countries, public demonstrations in Norway are characterized by differences in socioeconomic status. In order to assess the effect of social media on this aspect of participation, we need to establish the background profiles of different groups, those who take part in demonstrations after learning about them via the social media and those recruited via other channels. In Table 1 we compare the profiles of people who learned about the demonstrations through different channels, as expressed in answers to the question: ‘Where did you first receive information about the demonstration?’ Table 1 shows an analysis of the groups that received information through social media, personal contact, mainstream media and email/SMS respectively. We show both the most recent demonstration in which the respondent reported participation and the Rose Marches.
How did you first receive information about the demonstration?
Note: City dweller: Living in Oslo, Bergen, Trondheim or Stavanger.
The regression analysis shows the probability of any individual being recruited to participate in demonstrations through different channels. The results show that those mobilized through social media channels are indeed distinct from those mobilized via personal contacts and mainstream media. In addition to and controlled for young age, those at the lower end of the household income scale are more likely to be recruited to demonstrations through social media networks. Similar results are obtained after substituting alternative indicators of socioeconomic status for household income such as personal income (β −.192, p≤.01) and full-time employment (β −.881, p≤.001). In contrast, the highly educated are more likely to be recruited via personal contact, email or SMS while older respondents are more likely to be recruited through mainstream media.
With regard to the Rose Marches, the pattern is less pronounced. Age remains significant and negative, while the variables measuring socioeconomic status were insignificant. One interpretation of these results is linked to the enormous popularity of these demonstrations and the fact that Facebook was such a dominant medium for channeling information. On this occasion, the socioeconomic profile of Facebook as an information channel was arguably less pronounced because such a large proportion of the population used it as a means of communication.
The results in Table 1 are a strong indication that the use of social media does indeed serve to mobilize other groups than those that are mobilized through other channels. This in turn indicates that there is no digital divide in the civic usage of Facebook, rather the opposite. In what follows we will substantiate this proposition further by looking into the role of social media in terms of the individual and mobilizing agency dimensions of mobilization.
Social media and the individual agency dimension of mobilization
The individual agency dimension of mobilization concerns people’s motivation to participate and relies on access to information, self-efficacy, and interest (Verba et al., 1995). In order to assess the impact of the individual motivation dimension of mobilization we examined the role of Facebook in conveying information about demonstrations in general and about the Rose Marches in particular, and compared this with other channels of information. Using data on the Rose Marches we looked at a wider range of informational and expressive functions provided by Facebook and Twitter respectively.
The data collected in March/April 2011 allow us to assess the relative importance of different media in spreading information about public demonstrations in general, while the August survey included specific questions about the July 25 Rose Marches. As the results in Table 2 make clear, while Facebook is an important source of information about demonstrations in general, it was the leading source of information in relation to the Rose Marches.
How did you first receive information about the demonstration? Respondents who have taken part in public demonstrations.
In the case of the Rose Marches, 40% of those taking part first heard about the demonstration via Facebook. While the Rose Marches were extraordinary events, this recruitment pattern does not appear to be completely anomalous in comparison with other recent demonstrations and rallies. According to the April survey, 26% first heard about the last demonstration in which they participated via Facebook, 25% via the mass media (newspapers – print and online, TV, radio). Other digital media were less important – email accounted for 10%, other social media for 2%. Personal contact was also a major factor for participation in demonstrations more widely.
Facebook’s dominance in communicating information about the Rose Marches is not surprising given that the initiative was aired on Facebook to start with, and took off more or less spontaneously from there. In more general terms, we would like to argue, these findings show that whether we look at demonstrations and manifestations of public opinion in general or the Rose Marches in particular, Facebook emerges as a highly important means of mobilization in Norwegian society.
Breaking down how different social media users’ used specific social media following 22/7, we found that while Facebook played a major role, Twitter did too, for their respective user segments. Table 3 shows the uses of Facebook and Twitter during the events by people who used these applications.
Use of Facebook and Twitter for different purposes by age.
Notes: Percentage of Facebook and Twitter users responding ‘to a large degree’ or ‘to a very large degree’. Panel and population survey combined. Twitter users aged above 70 not shown in the table.
Table 3 reaffirms the pivotal role of Facebook as an infrastructure in mobilizing participation in the Rose Marches. The use of Facebook, moreover, was much more pronounced among the younger age groups. However, the table also shows that Facebook fulfilled a more expressive function for people of all ages; users posted messages of sympathy to the bereaved and used Facebook as a means of working through their own sense of grief and loss.
In general, younger social media users tend to rate Facebook and Twitter as the most important media on all counts. Aside from their capacity to mobilize populations, the most pronounced differences have to do with accessing and disseminating news and information, where Facebook, and even more so, Twitter, are generally more important to the younger users. Compared to Facebook, Twitter played a limited role in spreading information about the Rose Marches, but was a highly significant source of news about the terrorist attacks. So even though Twitter’s penetration in the population is lower than Facebook’s (12% as compared to 70%), it can play a significant role in galvanizing support for a public event, and, not least, in facilitating the dissemination of news. A particular feature of Twitter’s news propagation is the use of re-tweets, which allow users to record the importance they attach to an item of news. Popular tweets thus spread very quickly through cascades.
On the basis of our analysis, the impact of the social media was clearly significant at the level of individual agency in processes of civic and political mobilization. In the first instance, this impact is evident in relation to the motivational aspects of individual agency – that is, information, interest, and self-efficacy (Verba et al., 1995). With regard to information, our analysis shows that people first heard about the attacks predominantly via Facebook. Facebook also provided a space where people could post expressions of support in the period around the Rose Marches. The data collected in March/April on the general pattern of demonstrations confirm the importance of Facebook in spreading information about offline demonstrations in Norway in general.
With regards to the other key elements of individual agency, interest and self-efficacy, we should point out that information received through social media enables interactivity and observation of the actions of others. Our data show that social media fulfilled a number of purposes during and after the attacks. Users could express emotions and sympathy publicly, for instance, and it is reasonable to assume that the use of this social media inspired people to participate in public demonstrations. It is, for example, interesting to note that heavy social media users expressed the strongest sentiments about post-22/7 society; they also had a much stronger sense of ‘community and togetherness’, than the less active users did, although they were also more fearful and concerned about the possibility of more attacks in the future. 1 Social media may hence have amplified sentiments in the aftermath of 22/7, thus increasing individuals’ motivation to participate in organized events.
In light of our results, then, social media have changed how information about public demonstrations is spread, and indeed how people become motivated to participate in demonstrations. In the next section we turn to the mobilizing agency dimension of mobilization processes and explore the relationship between political and civic organizations and social media as channels for spreading information about demonstrations.
Do social media replace or complement established mobilizing agencies?
Theories of mobilizing agencies emphasize the importance of networks in getting information out to citizens and inducing them to participate in civic events (Rosenstone and Hansen, 2003). Traditionally, civil society organizations and political parties have been important structures in this context, because of the networks that they create through membership and affiliation. Civil society organizations and political parties set up information and communication structures adept at enabling rapid and efficient contact with their constituencies. Between organizations and members there exists, we assume, an amount of trust. This trust is what lends legitimacy in the eyes of the constituency to calls from the organization to rally behind some cause or other. Civil society organizations and political parties have in this sense specialized resources for mobilizing members and the public, because they can combine networks, legitimacy, and a set of shared normative expectations.
An important question when analyzing the impact of social media on processes of mobilization is whether they challenge, replace, or supplement established structures. As described above, social media’s affordances allow people to manage their networks through lists of friends. These networks are both structures of information and structures of support and legitimacy for different causes. In addition to person-to-person relations, some social media, such as Facebook, allow members to create ‘groups’ or ‘pages’. In the following we refer to both as ‘groups’ for convenience. 2 A Facebook group can be affiliated to an existing, offline organization or exist online only. These Facebook groups will presumably act as online organizations, propagating certain standpoints and linked to a network of followers. Given the network functionalities of Facebook, every member of a person’s network is able to observe how much that person supports a given Facebook group; information which can spread subsequently through group members’ own networks.
In what follows we examine the relative importance of different factors to participation in demonstrations. Table 4 depicts four different models, each assessing the impact of the background variables, of memberships in offline voluntary organizations and of political and civic Facebook groups. We focus specifically on offline organizations and Facebook groups with a political agenda, broadly conceived. The definition therefore includes social and humanitarian organizations and trade unions. The composite indices measuring number of discrete Facebook group memberships and organizational memberships were standardized in order to better compare effects. It should be noted that our focus here is on membership of voluntary organizations, Facebook groups, and participation in demonstrations in general. Only data from the April 2011 survey are used here, as the second wave of the survey omitted questions about Facebook group membership.
Participation in most recent demonstration (logistic regression).
Notes: Wald-tests are used to determine significance levels. * Significant at the 90% level. ** Significant at the 95% level. *** Significant at the 99% level.
In model 1 we assess the effect of different background variables on the likelihood of participating in at least one demonstration in the past six months. Having participated in a demonstration is negatively associated with age and positively associated with education. The model hence shows that the younger and more highly educated people are, the more likely they are to participate in demonstrations. Women likewise have a higher propensity than men. The positive effect of education is in line with other analyses of participation in demonstrations in Norway and in other European countries.
In model 2, we examine the relative impact of membership in offline voluntary organizations on participation in demonstration activities, after controlling for background variables. The analysis finds a strong, positive correlation between membership in political and philanthropic organizations and participation in demonstrations. In this analysis age remains significant, while education and gender lose significance.
Model 3 explores membership of political and Facebook protest groups and evaluates the predictive power of this variable on demonstration participation. As is shown, membership of offline political and philanthropic organizations and membership of one or more Facebook groups have strong and almost equal predictive power in relation to participation in offline demonstrations, when controlled for background variables. The impact of voluntary organization membership in model 2 is only slightly weakened when Facebook groups are introduced in model 3, and the explained variance increases substantially.
This analysis supports the argument that Facebook has an independent effect on mobilization, supplementing established civil society as a structure enabling and facilitating popular mobilization. When we look at different types of political and civil society organization and different types of Facebook group more specifically, as is done in model 4, it is clear the mobilizing effect of different types of membership varies greatly. Membership of a political organization other than a political party stands out as the strongest predictor of all, a category including most (new) social movements, for instance environmental, peace, human rights and feminist organizations. Membership of a trade union and social and humanitarian organizations also increases the probability of demonstration participation. Among Facebook groups, being a member of local and international protest groups, political parties and local politician groups also has an effect. Although the effects of membership of political parties and local politicians’ Facebook groups are rather weak, it is still interesting to note that affiliation with such organizations has an effect on Facebook, but not when linked to offline membership. Membership of other Facebook groups – that is, groups without any political purpose at all – is weakly, but significantly negatively related to demonstration participation.
So far we have established that Facebook groups constitute an independent addition to well-established civil society organizations as mobilizing agencies. It remains to be examined whether this structure attracts other groups than those represented by traditional organizational structures. As Table 5 shows, membership of politically relevant voluntary organizations and membership of Facebook groups are indeed accompanied by different sociodemographic profiles. Members of voluntary organizations, as column 1 shows, tend to be older, of the female sex, and better educated; household income is also higher. As shown in column 2, Facebook group members, in contrast, are younger and there is no evident over-representation of people in the high socioeconomic status bracket. On the contrary, the household income of political Facebook group members is significantly lower. It should be noted that substituting household income with an estimate of the size of the household income for each adult family member (in 100,000 Norwegian kroner) also yields significant coefficients (β –.023, p≤.05) (analyses not shown in table). In the analysis presented in column 3 we take account of the fact that some Facebook group memberships reflect offline organizational engagement, inasmuch as organization members often join an online corollary of the offline organization. After adding politically relevant organization memberships into the equation in an attempt to isolate the phenomenon of joining political Facebook groups as something other than a corollary of membership of an offline organization, an even sharper social profile emerges. In addition to low household income, low education is now also significant and the coefficients of low age and income are higher.
Politically relevant voluntary organization and Facebook group memberships. Logistic regression.
Notes: Wald-tests are used to determine significance levels. * Significant at the 90% level. ** Significant at the 95% level. *** Significant at the 99% level.
Thus, the analysis demonstrates two quite distinct circuits related to affiliation with offline organizations and Facebook groups. As pointed out above, organization memberships and Facebook memberships both increase the probability of demonstration participation. However, these ‘institutions’ affiliate groups which have diametrically opposite characteristics. While members of traditional organizations tend to do better in socioeconomic and educational terms, and are older, members of political Facebook groups tend to have lower socioeconomic status, are less well educated and are younger.
Based on our analysis, we can therefore conclude that Facebook constitutes an addition to conventional organizations as a mobilizing structure – at least in terms of demonstration participation, and that the predictive power of voluntary organization membership and Facebook group membership respectively is comparable.
Discussion and conclusion
In this paper we have analyzed a set of questions related to the role of social media in civic and political mobilization processes. Based on a view of mobilization that combines individual-level characteristics (individual agency) with a structural view (mobilizing agency), we have asked what changes the emergence of social media has entailed at these two levels. The results indicate quite profound individual- and structural-level changes; social media mobilize specific sociodemographic segments, they change how individuals are informed and motivated to participate and they constitute a new form of mobilizing agency that neither simply reflects nor crowds out existing formalized and established structures. Participation in Facebook groups has a strong and independent effect on mobilization, and social media such as Facebook must therefore be conceived as supplements both to the organizational establishment and to mainstream media as information structures facilitating mobilization.
We need to know, however, whether these changes to individual and mobilizing agency affect the scope of mobilization and hence serve to reduce socioeconomic differences. If so, it would indicate a lack of digital divide in the propensity to use social media for civic purposes such as mobilizing people to join offline civic events. We do find in our analysis significant differences between those who join demonstrations because of information conveyed by social media and those who are encouraged by established civil society and political organizations. Participants mobilized through social media are characterized by lower socioeconomic status and younger age than those mobilized through established civil society organizations. If we compare with mobilization effected by mainstream and other media, there is a similar pattern: social media recruits are less well paid, they are younger and less likely to have a full-time job than the rest of the population.
So does mobilization through social media lead to the inclusion of new groups that were previously under-represented in civic and political demonstrations? This question cannot be answered on the basis of the present data, since they do not capture developments over time. Still, it is an interesting point that the socioeconomic profile of participants recruited through social media differs from what has been shown to dominate in earlier studies of political mobilization in Norway. According to these studies, there is a correlation between education and the propensity to participate in demonstrations (Pettersen and Rose, 1996). Social media therefore seems to offer a channel that supplements established political and civil society organizations, by reaching different and less privileged groups.
Based on our findings we therefore argue that we might be seeing the early stages of a transformation of civic and political means of mobilization. Social media seem to represent an alternative structure alongside mainstream media and established political and civil society organizations in that they recruit members in different ways and from different segments of the population, thus conforming to the mobilization thesis (Dahlberg and Siapera, 2007; Dahlgren, 2005). Earlier studies of internet use and political participation indicate that the internet may have a different effect on the young and the adult population – in other words, a mobilization effect on the young and a normalization effect on the adults (Hirzalla et al., 2011). The present study indicates that this mobilizing effect may not be limited to a given generation or age group, but have a transformative potential linked to differences in socioeconomic resources.
Insofar as this study reaches a different conclusion than several older population-based studies (Chadwick, 2006; Margolis and Resnick, 2000; Norris, 2001), it would be interesting to ask why. One factor, we believe, is linked to advances in technology since these studies were carried out. The affordances of social media and Web 2.0 are fundamentally different from the affordances of Web 1.0 and the internet in general. Collaborative democracy through social media is enabled by a combination of technical architecture (the social media affordances), network structure and effects, and cooperation-enabling social norms that are widely shared by social media users. The technical architectures and social dynamics of information that characterize the production and exchange of information through social media are modular, flexible, mobile, and decentralized. They allow many people to act in conjunction and to coordinate actions which cohere and aggregate, by way of information cascades across digital networks, into mass mobilization processes or protests which manifest themselves online and in the real world.
A second distinguishing factor in our study is the level of case and time specificity of the data used. As pointed out by Hirzalla et al. (2011), one reason why survey-based analyses of political participation have found a normalization effect rather than a mobilization effect, may be because they were unable to capture the time and specific nature of political participation. In the present study, we asked respondents a series of very specific questions about the online/offline process leading up to their participation in demonstrations, and we also specified in these questions which demonstrations respondents had taken part in in a given period. The data obtained by the survey on participation in the Rose Marches were even more specific and detailed, because in this case it was possible to ask all respondents how they were mobilized to join a particular and identified set of demonstrations.
The examination of offline demonstrations and their links to online activity is, we believe, of particular value to the study of the mechanisms involved in the transformation of political participation in the Web 2.0 era. One reason, which is related to the point above, is that because the demonstrations are marked points in time, case- and time-specific analyses can be performed on the basis of population-based surveys. We agree with Carpentier and Dahlgren (2011) on the need to specify and differentiate the term ‘participation’ when studying the transformation of democracies in the digital age. Offline demonstrations represent a strong form of civic and political participation because they are so clearly linked to action. They therefore provide an ideal setting for examining the mobilizing potential of social media and comparing it with the efficacy of other mobilizing agencies.
Footnotes
Funding
This research has been funded by the Norwegian Research Council, through the research program VERDIKT.