Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Political communication has a central role in every strategy that seeks access to, or the maintenance of, power. For that reason it is not surprising that in democracies the electoral campaign is the most intense political period of communicative activity (Martínez Nicolás 2007:211). Especially in electoral campaigns, political parties multiply their number of communicators, diversify their channels of communication, and intensify their efforts in order to get their message across to the greatest possible number of receivers, in a marketing effort (Petrocik 1996).

This pattern of political behavior is also repeated in the digital environment. The political parties have found in social networks a new and privileged scenario for their electoral campaign (Anduiza et al. 2010). For the candidates, the social networks also represent a communicative resource of high value as a consequence of its interactive potential, and its philosophy of direct and personal relationships.

However, exporting campaigning activities to the digital arena of the social networks does not only entail extra-marketing efforts for parties and candidates. It also involves potential changes in the focus of these energies. In fact, this new way of campaigning is what Zittel has conceptualized as individualized campaigning: the use of the Internet to promote the personal dimension of the electoral message and to promote the proximity of the candidate to the elector, increasing his or her capacity of personal interaction and dialogue with the community (Zittel 2007:6–7; Zittel 2009).

Studies such as those of Gibson (2010), Gibson and McAllister (2011), Gueorguieva (2008), Gulati and Williams (2010), Steger et al. (2010) among others, have examined the importance of the social networks and tools of the Web 2.0 (YouTube, MySpace, Facebook, etc.) in the most recent electoral campaigns held in developed democracies. One conclusion that these studies have in common is the fact that there is a growing tendency to use the social networks as a channel of political communication. Nevertheless, little evidence exists about these phenomena beyond this verification. The behavior that parties and candidates show on the social networks; the degree of coherency among them; and the principles of the so-called Web 2.0, are questions that have still barely been addressed in the academic literature (Bode et al. 2011; Golbeck et al. 2010; Lassen and Brown 2010; Gaines et al. 2009; Williams and Gulati 2009, 2010; Yannas et al. 2011). The role of social media in different spheres of electoral political democracies requires greater understanding.

This study seeks to contribute to the accumulated knowledge about electoral campaigns and social networks by means of an empirical study that employs the case study as a methodology. For that reason it focuses on the impact of Twitter on the electoral campaign of the local Spanish elections of May 2011. On the basis of this case study, the following questions are addressed: Which candidates to the mayoralty employed Twitter during their campaign? To what extent did they use Twitter? Did they manage to widen or deepen their network during the campaign? Did they really use Twitter to boost their dialogue with the electors linked to this network? Did the candidates’ gender, age, political party, municipality affect this behavior?

This study is presented in six sections. Following this introduction a second section is dedicated to contextualizing the electoral campaign developed in Spain in the recent local elections, justifying at the same time the political opportunity to employ Twitter as a channel of personal interaction with the electorate. The third section identifies the analytical strategy followed in order to address the research questions of this study. The fourth section presents the results of the analysis undertaken. The fifth section discussed the main findings. The paper closes with a sixth section that offers the conclusions of the study.

2 The 2011 Local Elections in Spain. Why “individualized campaigning”? and Why Use Twitter To Do So?

Every party’s electoral campaign is based on a communication plan of a persuasive character. This program of persuasion is generally oriented toward mobilizing and attracting the vote of three types of electors (Holborok 1996; Scher 1997; Barnes 2005). The first group is that of the ‘party faithful’, which represents the captive or base vote of the organization. The second group is that of the voters who are susceptible to vary the orientation of their vote between one electoral campaign and the next. These ‘non party’ electors (Dalton 2002) are the source of the fluctuating or volatile vote, and represent the opportunity for parties to add a strategic percentage of additional votes to their base vote. The third group is that of the electors who generally do not vote.

The effort that the parties dedicate to persuade each of these groups of electors varies in function of the structural and circumstantial conditions of the electoral competition (Vanaclocha 2005). In political contexts where the structural form of the electorate is strongly ideologized and/or intensely identified with a party, the parties dedicate a large part of their resources to mobilizing their ‘party faithful’. For that reason they employ campaigns that have a strong ideological or identity bias. On the other hand, in contexts where there is a significant percentage of floating and/or abstaining voters, this becomes a second focus for priority attention. The strategic value of this group’s vote achieves the highest level when, circumstantially, the elections are perceived as elections of rupture or change. In these cases, the parties develop campaigns of persuasion that are more centered on the candidates and on their specific projects than on the ideology and identity of the organization.

In Spain party ideology and identity represent determinant factors of the orientation of the vote for the majority of electors. Nevertheless, Spaniards perceive the municipal elections as elections of second order. For this reason in the local political arena they present electoral behavior that is distinct to that demonstrated at national level elections. The particularity of Spanish electoral behavior at municipal level resides in two tendencies: the abstention differential and the dual vote (Molins and Pardos 2005; Montabes Pereira and Ortega Villodres 2011). On the one hand, the municipal elections mobilize the electors to a lesser extent than the general elections. On the other, those electors who go to the polls in the municipal elections tend to consider dimensions of the electoral offer that go beyond ideology or the brand of the party. Among these alternative dimensions may be highlighted the qualities of the candidate for mayor–those who the electors feel closest to, as politicians–and their projects of city management–by those who feel directly affected by them (Martínez Fuentes and Ortega Villodres 2010a).

The Spanish electoral arena that presents the highest rate of floating voter and abstentionism is at local level. Consequently, it is at this stage where these types of electors receive the highest level of attention from the parties. In general terms, the large parties encourage their local groups to run an electoral campaign that sells not only the brand of the party, but also the figure of their mayoral candidate; specifically their personal identification with the local community and their concern with specific problems. As a result, in Spain, the design and implementation of the municipal electoral campaigns has a strongly personalized or presidential nature (Natera 2001; Martínez 2008; Martínez Fuentes and Ortega Villodres 2010b; Criado and Martínez Fuentes 2010, 2011).

In particular, the municipal elections of May 2011 opened up a scenario, for the largest Spanish parties, that especially favored the design of strategies of electoral competition orientated to the floating voter and the abstentionist elector by campaigns that were centered principally on the promotion of candidates to the mayoralty. Why? In first place, the party brand was not particularly attractive on this occasion. In second place, the municipal competitions were not perceived as ‘second order’ elections. On the other hand, all the preelectoral polls coincided in the prediction that a significant number of citizens were going to punish with their vote, or abstention, both the party who govern the nation (The Socialist Workers’s Party–PSOE, in their Spanish initials) and their President. This is why this election was seen as ‘primaries’ for the following general elections. In sum, the 2011 municipal elections were viewed as elections of change.

In the large Spanish cities, in particular, these local groups decided to adapt to the circumstances by adopting the technique of “individualized campaigning” as a campaign strategy. To put this campaign model into practice, a significant number of candidates looked to the ‘online’ social networks. The existing high degree of social receptivity in Spain to the political use of the social networks helps to explain this decision. This was highlighted by an opinion survey undertaken by Intelligence Compass (2010) months before the holding of the elections. 96 % of Spanish politicians who used social networks were found to consider them to be channels of either ‘significant’ or ‘outstanding’ importance in their contact strategy with citizens. In parallel, 86 % of Spanish citizens surveyed stated that politicians should use social networks to maintain themselves in contact with electors.

Three fundamental reasons determined the choice of Twitter as a key social network for this election campaign. The first is the degree of diffusion achieved by this social network. In Spain, more than 60 % of the population are users of the Internet (EUROSTAT 2011). In addition 83 % of Spanish Internet users employ social networks and 25 % subscribe to Twitter. The second is its capacity to facilitate access to nonparty and abstentionist electors. The majority of Internet surfers and users of social networks are concentrated in urban locations where there are larger numbers of young people with higher levels of formal education (INE 2011). In fact, 90 % of young Spaniards aged between 16 and 20 are linked to a social network (Fundación Orange 2011). The third argument is the potential associated with Twitter to endow the campaign with a personal, direct, interactive and speedy style. According to a study undertaken by Cocktail Analysis (2011) among the Spanish users of social networks, many who were surveyed showed a preference for Twitter because of two specific attributes: its speed and direct character. For these reasons, recent comparative studies, such as Hanna et al. (2011), Jaeger et al. (2010), Pole and Xenos (2011), and Tumasjan et al. (2011), have emphasized the use of this resource for electoral purposes.

3 Analytical Strategy: Inquiring into Who Tweets?, How Much They Tweet?, and What for?

This section provides the analytical strategy that has guided this study. The objective of our research is to reveal two key facets of the use of Twitter during the electoral campaign of the municipal elections held in Spain in May 2011: namely, the presence and behavior of the candidates in this social network. To achieve this stated goal we raise the following research questions: How many candidates used Twitter during the campaign? Did gender, age, political party, and size of the local community affect this behavior? To what extent did they use Twitter? Did they manage to broaden or deepen their network throughout the campaign? Did they really use Twitter to promote their interaction and dialogue with electors who are linked to this network? We designed the following research strategy with the aim of finding responses to these questions.

In first place, we considered the size of our universe of study. The mayoral posts of 8,808 municipalities were under competition in the May 2011 Spanish local elections. In consequence, we decided to focus our attention on a specific segment of that group. In this way, we decided to limit our object of study to the mayors of large cities who sought to renew their posts by competing as candidates for the mayoralty in the 2011 elections.

The decision to study only mayoral-candidates derives from the fact the local Spanish mayoral political system, classifiable under the formula “strong-mayor-form” (Mauritzen and Svara 2002), provides the mayor with a high level of personal political visibility and influence. Assuming that the use of Twitter in an electoral campaign seeks to exploit the personal facets of the candidate, we understand that the candidates that are at the same time mayors were the best known.

The idea of studying these types of actors, only in the context of large cities, is justified by one reason. Given that in Spain the use of social networks is concentrated in the large urban areas, the mayors who achieved most electoral returns through the use of social networks would be precisely those who led in large cities.

To identify the mayoral-candidates from large Spanish cities who are users of Twitter in electoral campaigns, we decided to track which of them had an official Twitter account on the date of the start of the electoral campaign (6th of May). In this tracking and identification process, we decided not to discriminate if those accounts were managed by the candidate or by personnel from their communication team.

Besides, in order to clarify if determining factors exist in the use of Twitter between mayoral-candidates of large cities, we opted to observe their own attributes and those of their environment. Therefore, we took into consideration the following factors:

  • Candidate gender: the general statistics indicate that the users of Twitter are mostly men (National Observatory of Telecommunications and Information Society 2011).

  • Age: the general statistics highlight the fact that users of Twitter are mostly young (National Observatory of Telecommunications and Information Society 2011).

  • Party to which they pertain: the candidates of the PSOE may have greater incentives to emphasize the personal component of their campaign by using individualized campaigning, due to the crisis of popularity of the PSOE and of their national leader.

  • Size of their community: the mayoral-candidates of the most populous cities show a greater willingness to use Twitter in their campaigns, as a means of getting closer to their citizens.

Then, we measured the intensity of use that the mayoral-candidates of large cities made of Twitter during the electoral campaign. For that reason we identified the number of tweets emitted by these candidates between the 6th and 20th May.

Next, we evaluated the density of the network of the mayoral-candidates on Twitter and the effect that the campaign had on it. In first place, we identified the number of profiles that the candidates had as followers at the start and at the end of the campaign. In the same way we identified the number of profiles that the mayoral-candidates followed in Twitter.

Finally, we analyzed if the use the candidates made of the Twitter network in their electoral campaign was coherent with the interactive and conversational logic of the 2.0 philosophy. For that reason we counted the number of tweets sent by the candidate that involved a direct response to a user in this social network. (That is, we identified how many tweets appeared with the ‘dialogue balloon’ symbol.)

4 Results

In this section, the data collected during the fieldwork and its statistical treatment is presented in two separate parts. In the first we discuss the use of Twitter in the electoral campaign. In the second we address the type of use that this network has received, and identify its determinants, throughout the campaign.

4.1 Who is Using Twitter as an Instrument for Campaigning?

Table 14.1 shows the initial universe of observed cases and the final universe of the cases studied. In the Spanish municipal elections of 2011, 55 mayors of large cities sought to renew their candidature for the post. After tracking their presence in the Twitter network, we concluded that 39 of them had an operative account at the start of the electoral campaign (around 70 %). Considering that in the previous Spanish municipal elections held in 2007 there was no electoral use of Twitter, our data allowed us to confirm that Twitter is a tool that has a new and high degree of popularity among local Spanish political leaders, at least during the last election campaign.

Table 14.1 Mayoral-candidates using Twitter during the last electoral campaign
Full size table

4.2 Factors Explaining the Extent and Style of Twitter Use in the Campaign

Do gender, age, party affiliation, and size of the municipality of the candidate, may explain the use of Twitter as a channel of communication in the electoral campaign and its means of employment?

Table 14.2 shows that there were significant differences between men and women in the level of utilization of Twitter during the electoral period under analysis. While 74 % of male candidates had a Twitter account, only 58 % of female candidates made use of this network. Then, these data reflect general population statistics for male/female use of Twitter in Spain.

Table 14.2 The use of Twitter by gender
Full size table

Gender differences can also be seen in male and female candidates’ use of Twitter during the campaign. The average number of messages sent is greater in the case of the men. From its part, the volume and evolution of the candidates’ virtual communities also show differences on the gender variable. Men have greater success in adding supporters. They manage to increase the size of their community on the network by 25.11 % on average during the campaign. Additionally, men also pay more attention to the work of listening; thereby more rapidly increasing the quantity of people who follow them (13.51 % in the case of men, against 2.04 % for women).

On the other hand, these data result quite intriguing if we contrast them with the fact that there is no a significant difference between male and female candidates behavior when the volume of ‘dialogue’ is analyzed (16. 18 % in the case of men, against 15.55 % for women). Since the level of dialogue sustained by female and male candidates cannot explain why male candidates were more successful in adding supporters, opening new lines of research within this framework would be needed to clarify the reasons of this manifestation.

Regarding the independent variable of age, Table 14.3 shows that the candidates who are aged between 35 and 44 have a greater relative presence on Twitter. This data, however, does not allow us to establish a direct relation between the youth of the candidate and the propensity to use Twitter as a channel of communication in the electoral campaign. As can be seen in the table, the second group who have a greater rate of presence on Twitter are candidates who are older than 65. Then, we can conclude that there is no correspondence between our data and general statistics about young/elder population use of Twitter in Spain.

Table 14.3 Data by age
Full size table

The age of the candidates does not appear to clearly explain level of activity on Twitter. Although the level of activity of those older than 65 is very much lower than that of candidates from the younger age groups, there is no growing activity on Twitter as age groups descend. The groups of 35–44 and 55–64 years old show slightly above the average number of messages, while this effect weakens in the age group between 45 and 54. Therefore it might be suggested that age does not determine the development of a specific level of activity in the digital age.

The explanatory capacity of age increases for the dependent variables ‘differential of monitoring’ and ‘differential of supporters’ of the candidates between the start and end of the electoral campaign. The youngest age group has patterns of behavior that are very distinct to the rest. In this group, while their number of people followed increased to 34.35 % during the whole campaign, the people who decided to follow them increased to 49.13 %. The results show that this group produces more increments in both categories, and is also unique in having greater growth in the number of people who listen to them. In the remaining groups the growth in followers is much less (between 18.8 and 25.03 %). It is worth emphasizing the data that shows a decrease in the differential in the age group between 55 and 64 years old (−9.11 %). In the other two groups the increases are marginal and barely reach 3 %.

These data invite to take into consideration two possible explanations about the relative success of the youngest candidates. One of them is that they were more successful in adding supporters because they sustained a higher level of dialogue. An alternative one has sociological roots: their success responds to the mere fact they were young–being the candidate’s youth considered as an element of identification between the candidate and the majority of the Twitter users.

The third independent variable that is analyzed in this study is the political party to which candidates belong. Of the 39 cases that comprised our final universe of study, 38 correspond to candidates who are members of the Spanish Socialist Worker’s Party (PSOE) or the Popular Party (PP). From now on we focus on this double category PSOE-PP.

In this respect, Table 14.4 makes it clear that defending the political colors of the PSOE or of the PP does not affect use of individualized campaigning via Twitter. That is, the candidates of the PSOE and the PP had a similar predisposition to have a presence on Twitter in this electoral campaign (74 % PSOE against 75 % PP). However, to pertain to one or the other party does appear to affect the way in which Twitter was used during the election.

Table 14.4 Division by party
Full size table

The average number of tweets employed by the candidates of the socialist party was greater than that of the PP candidates. A more marked difference between both groups of candidates was found in the attitude toward dialogue that they adopted in Twitter. In this sense the percentage of dialogue interactions of the PP with the electorate was double that of the PSOE. Furthermore, one can observe large differences in the evolution of the networks of followers of the candidates of both parties. The increase in the network community of the candidates of the PP was also greater than that of the candidates of the PSOE (27.43 % against 15.36 %). The advantage of the PP was also notable in relation to the ‘followers of candidates’ differential. While the number of accounts followed by the socialist candidates increased throughout the campaign by an average of 3.02 %, in the case of the PP the differential ascended to 22.04 %.

These data are reasonable in electoral terms. The concrete electoral context of demobilization of socialist voters gave the socialist candidates more incentives to make strongest marketing efforts in their management of the social network. Then, they posted more tweets than the popular candidates. However, the political pressure over the socialist party made socialist candidates present themselves more active than interactive in the social network. Popular candidates behaved in the opposite way. They focused on follow new people and dialogue with the electorate, achieving in this way better results in terms of increments in the number of followers.

The last independent variable analyzed in this study relates to the size of the municipality. According to Table 14.5, the larger the size of the municipality of the candidate, the greater his/her propensity to use Twitter as a channel of communication with the electorate. The mayors of the municipalities with more than 200,000 inhabitants have a rate of participation in this social network of 67 %, while those who compete in smaller municipalities have 7 % points more (74 %).

Table 14.5 Division by size of municipality
Full size table

These data seem to be coherent. The bigger the city, the fewer chances the candidates have to interact personally with the electorate and the more useful the social networks may be to do so. In a parallel logic: the smaller the city, the fewer citizens are active in social networks, so the fewer incentives the candidates have to make extra-marketing efforts in the digital environment.

If it can be said that the candidates of the municipalities of less than 200,000 were more active in the use of the network (they made more tweets), it is true that they also had less dialogue with their followers in their virtual community. In fact, the mayors of municipalities of more than 200,000 inhabitants posted a percentage of dialogue messages that was double that of their counterparts of municipalities of lesser size. Other data equally shows that the mayors of larger municipalities took better advantage of the possibilities of Twitter in terms of their network of people followed, and people who were following them. The following activity of the mayors of bigger municipalities increased to 27.51 % during the electoral campaign, against a 19.22 % increase experienced by the number of people followed by the mayors of municipalities with fewer inhabitants. The difference is still greater if we consider that the followers of the candidates of the larger cities increased to 17.81 %, against 3.69 % in the case of the candidates of the less populated cities.

We find a double and basic explanation for these facts. Since in bigger cities the social network is denser, the mayoral candidates in these environments had potentially more opportunities to increase their number of followers. Moreover, since in bigger cities mayoral candidates finally demonstrated themselves more prone to dialogue, they were in fact more attractive to followers.

5 Discussion

In this section we emphasize our main research findings. The five points selected have been chosen for their descriptive/explanatory value, and also for their potential to inspire new lines of research.

Although in 2011 the PSOE had greater incentives than the PP to run a local campaign based on the strategy of individualized campaigning, in the large Spanish cities both parties similarly decided that their greatest electoral attraction were their candidates. Then PP and PSOE exported their campaigning activities to the digital arena of Twitter, encouraging their respective mayoral candidates to make extra-marketing efforts in this environment. This fact reveals that the main and major Spanish political parties contribute to the promotion of a visible tendency toward the personalization of local elections. That also invites us to suggest that this party behavior reinforces the already accentuated personalist character of local political leadership in Spain.

In second place, for the Spanish political elites the local political scenario has been a ‘test laboratory’ of the application of the technique of individualized campaigning through Twitter. This communication tool has achieved notable diffusion among the main candidates to the mayoralty of large Spanish cities. It still remains to be seen to what point the circumstances that surrounded these elections determined the high level of utilization’ of Twitter on this occasion. As a result, these findings raise new questions. Is this level of Twitter use an isolated case or will its use continue to increase in municipal and/or national elections that are less marked by the expectation of political change?

The socialist and popular candidates to the mayoralty of the large Spanish cities use Twitter to raise awareness about their electoral message, but also to increase their community of contacts and support through four techniques (namely: broadcasting tweets, following others on the network, ensuring that they are followed, and initiating dialogues with their interlocutors). Nevertheless, they do not all know how to exploit the entirety of the communicative potential that Twitter offers. Those who most strategically employed this tool included: the PP candidates, men, the youngest, and those from the most populous cities.

Without doubt the greatest novelty introduced by the use of Twitter in this electoral campaign was the possibility of establishing direct and fluid dialogues between candidates and electors. Taken together, the data from this study suggests that this possibility was not exhaustively exploited by most candidates. As a whole, the majority of the candidates’ interventions on the network were made to broadcast their own messages, and only a small percentage responded to messages of their interlocutors. This reality can be interpreted in two different ways. On the one hand, it might be suggested that the candidates to the mayoralty of the large Spanish cities, in general terms, are still at an early stage in their use of the social network, and for that reason did not know how to make best use of it. On the other hand, it can equally be argued that this process will demand the time, effort, and training of those candidates who are accustomed to a style of unidirectional campaign that is supported by the party, and that focuses on traditional means of communication.

Finally, now we do know that in Spain most mayoral candidates made new marketing efforts to develop their local campaign in Twitter. Throughout this study we have found as well preliminary data to approach the understanding of how these efforts were made and what is their impact on election politics. On the one hand, our work shows that efforts were made in a heterogeneous way, since the candidates’ incentives were varied and conditioned by different factors. On the other hand, our study indicates that these dissimilar efforts had an impact in the Spanish model of personalization of local politics. When candidates wished to learn and really exploit the communicative opportunities that social networks offered to them, leadership was still in force but the accent was not posed just in the leaders figure. The emerging way of personalization of Spanish local politics promoted by the best use of social network in the conduction of individualized campaigning is based on the bidirectional relation of influence between leaders and followers, mayoral candidate and electors. That is real leadership, is not it?

6 Conclusion

In this chapter, the importance of social media in electoral campaigns has been made clear. For that reason, an initial discussion has explored the relation between political communication and social networks during the time of electoral campaigns, and has highlighted the scarcity of empirical evidence about this phenomenon. Immediately following this analysis, the frame of reference in this case study has been specified. In this way, the presence and the behavior of the mayoral-candidates on Twitter during the recent local election campaign in Spain has been analyzed. This research design is both innovative and unique in Spain. Nevertheless, we consider that its interest transcends the Spanish academic world, given the scarcity of equivalent empirical work at comparative level.

The study has provided statistical results that have allowed us to respond to the questions that were posed at the outset. It has been made clear that Twitter has been used in an electoral campaign by the immense majority of the observed candidates. Equally, the existence of different patterns of use of the network as a channel of electoral political communication has been emphasized. Not all the candidates with a presence on Twitter have exploited the personal and interactive potential of this network in a similar way, nor have they been capable of making their network of support denser by fully employing the resources that it offers. Through investigating the incidence of various factors in the expansion and ‘style of use’ of Twitter as a campaign tool, we have revealed that gender, size of the population of the candidate’s municipality, and party membership seem to have an effect in the behavior on the mayoral candidates in the social network. However, just the two first variables seem to influence the dependent variable “presence of the candidates in Twitter”.

These results, taken together, have a merely exploratory value. A nuanced statistical treatment of these data is needed to reveal the insights of the facts they suggest. The consideration of other independent variables is equally required for further steps of research. Nevertheless, the interest of the data already presented is not small, given that they offer some initial clues from which to start investigating the phenomena of the virtualization of electoral campaigns, be it in the study of isolated cases or in comparative studies of different cases.

Therefore, it is expected that future studies can address some of the questions that have been left open by this chapter. Is the use of Twitter in campaigns a passing fashion or will it become consolidated as an electoral strategy of political communication? In this sense, it would be interesting to see the results of various consecutive electoral contests and to monitor this phenomenon over time. Is the use of Twitter more common in local or national election campaigns? Equally valuable would be a comparative analysis between electoral scenarios at distinct levels. What is the role of other social networks in the electoral campaign? We argue that research projects that are more ambitious, and that contrast the electoral use of other social networks such as Facebook, will greatly enrich accumulated knowledge about this question.