Is the voter only a tweet away? Micro-blogging during the 2009 European Parliament election campaign in the Netherlands
First Monday

Is the voter only a tweet away? Micro-blogging during the 2009 European Parliament election campaign in the Netherlands by Maurice Vergeer, Liesbeth Hermans, and Steven Sams



Abstract
This study explores the use of Twitter by candidates, in particular their networking and micro–blogging activities in the election campaign for the European Parliament elections of 2009 in the Netherlands. The main focus is on identifying what political aspects (e.g., party characteristics and candidate characteristics) influences their use of Twitter as a campaign tool. Furthermore, we explore the effectiveness of candidates’ activities on Twitter in gaining votes.

Contents

Introduction
Politicians’ use of Web 2.0
Hypotheses
Methods
Results
Discussion and conclusions

 


 

Introduction

In many Western countries, politics increasingly suffers from a decline of interest and participation in the political processes (Flickinger and Studlar, 2007). This not only applies to national politics but also, and even more so, to European politics. The European Union in particular suffers from a severe democratic deficit (cf., van Os, et al., 2007). In general, support for the European Union is low: only a small majority (52 percent) supports its country’s EU membership (European Commission, 2009). The steady decline of voter turnout (43 percent for the entire EU in 2009; European Parliament, 2009) for the European Parliament elections underlines this. The Netherlands, although its population supports the European Union membership quite strongly (78 percent), shows a voter turnout of only 37 percent, which is well below the EU average. Even though support for the EU remains relatively stable, it is not clear whether the steady decline in voter turnout is reversible.

Thus politicians are challenged to reduce this growing gap between citizens and politics. E–campaigning and in particular Twitter, which is the focus of this study, creates greater visibility and increased interactivity between politicians and citizens, and shows potential to mobilize people to become politically involved, which in turn should reduce the gap between politics and citizens.

The Internet as a campaigning tool

From the moment it became popular among the general population the Internet, in particular the Web, was viewed as a means of trying to reverse the declining political participation (European Parliament, 2009). Parties and candidates indeed increasingly use Web sites to communicate and connect with the electorate (Kluver, et al., 2007). This new campaign strategy — emerging at the end of the last century — was coined ‘professional’ by Gibson and Römmele (2001). With the rise of individualization and modernization in society and declining political involvement and interest in politics, parties search for new options to reach individual voters by using the Internet, and applying a marketing approach to target specific groups of people.

Even though studies on Web campaigning are numerous, most online campaigns analyzed are Web 1.0 campaigns (cf., Foot, et al., 2007; Norris, 2001, 2003; Tedesco, 2004). The concept of Web 1.0 indicates that Web sites are predominantly hierarchical and disseminating, from the politician and party directly to the citizens, using standard technology (predominantly html). The benefits of Web 1.0 in political campaigning have been described by Barber, et al. (1997): interactivity, potential for horizontal and lateral communication; non–hierarchical communication; low costs for users; speed of communication; no national or geographical boundaries; freedom from intrusion and monitoring by government. Although these characteristics are valid for Web 1.0, these options were still underutilized. Technical limitations and low user–friendliness still limited the extensive use by producers and consumers. Web 2.0 (Cormode and Krishnamurthy, 2008; O’Reilly, 2005), characterized by innovations (e.g., AJAX, the Web as a platform), is suitable for people to engage directly and interact with others on the Web. Keywords associated with Web 2.0 are bottom–up approach, sharing of content, online collaborating between people and socializing among people and networking. Today, in election campaigns, Web 2.0 applications are considered to be even more appropriate to close or, at least, decrease the gap between politics, politicians and citizens. Although the arguments in favor of Web 2.0 seem compelling the next section will show that the utilization of Web 2.0 in politics does not live up to its democratization and leveling promises.

The present study will focus on the candidate’s use of the micro–blogging and social networking service Twitter as a new tool for campaigning. Nevertheless, we stress that, even though academic attention is increasingly focused on Web 2.0 campaigning, studying Web 1.0 campaigning still shows theoretical and empirical refinements, such as longitudinal analysis (cf., Larsson, 2011), and cross–national comparative analysis (cf., Vergeer, et al., forthcoming).

 

++++++++++

Politicians’ use of Web 2.0

When reviewing the opportunities these new Web 2.0 technologies offer, we see that the architecture of Web 2.0 allows non–experts to use the Web and to contribute to it in a way that was not possible in the Web 1.0 era. This not only gives the possibility of closing the gap between politicians and the electorate, it also holds out the promise of to closing the digital divide between people in general. As such, it has potential to increase democratization and emancipation, especially for those categories of people caught in disadvantaged positions. Even politicians, especially those who receive little attention from the traditional media, can now publish their opinions easily through new and additional Web channels such as personal Web sites, (micro–)blogging sites and social networking sites.

With the introduction of Web 2.0, many parties, politicians and candidates adopted blogging, social network sites and sharing sites (De Zúñiga, et al., 2009; Gueorguieva, 2008), thus the question arises whether this is a new campaigning style, different from the other types. If so, what are the basic characteristics of this new campaigning style? Web 2.0, with popular examples such as Facebook, YouTube and Twitter, also allows politicians to individualize and personalize their campaigning style. By doing so, politicians try to decrease the psychological distance between themselves and voters (Caprara, et al., 1999).

An important benefit of the Web and (micro–)blogging in particular is that by ignoring the institutionalized traditional media it allows for greater autonomy of the candidate due to more direct communication. Using the Web also allows politicians and candidates to operate relatively autonomously from the party. It can be argued that some parties (e.g., right–wing parties) allow their candidates more autonomy than other parties (left–wing and conservative parties). Whether this is beneficial to the party’s strategy is unclear. It is very easy for candidates to debate online without the necessary restraint and so permitting slips of the tongue causing the party more harm than good.

A formalization of the prior arguments, which can be described as an e–optimist view, is the innovation hypothesis (cf., Schweitzer, 2008) which states that certain characteristics of new media technologies fundamentally change the way politics is brought to the public. It differs from the offline patterns of electioneering and offers opportunities to revitalize rational ideals on democratic discourse.

Although these are compelling arguments underlining the benefits of Web 2.0 for political engagement, the counter arguments are also convincing: those parties that are already in an position of advantage, established parties and those in power, will possess more strategic knowledge, but also have more resources to support the utilization of new technology to further advance their lead: through long experience these parties have built up more means, such as knowledge about campaigning, financial resources and human capital, than new and small parties. Margolis, et al. (1999) conclude in their study of the use of Web 2.0, that for the U.S. “politics as usual” prevails, whereas in the Britain there are also indications supporting the innovation hypothesis. The same holds for the use of a typical Web 2.0 platform (YouTube) by U.S. congressional candidates, that was augmenting — not replacing — traditional campaigning and was utilized by the best financed candidates (Gulati and Williams, 2010).

The two hypotheses — innovation and normalization — are mutually exclusive: if one hypothesis is supported by the data, the other one needs to be refuted. However, the real test of either of these hypotheses has yet to be conducted, because many studies have suffered one or more of four methodological constraints limiting rigorous testing: (1) data are often cross–sectional instead of longitudinal; (2) evidence originates mainly from the U.S.; (3) focused on candidate Web sites; and, (4) focusing on Web features and seldom on the content of communications (Schweitzer, 2008). In her longitudinal study of political Web sites in German elections (2002–2005), Schweitzer (2008) concludes that the innovation hypothesis was partially confirmed, as such, only for the parliamentary parties studied [1]. In contrast, the non-parliamentary parties have underutilized the potential of the Internet, due to limited financial and human resources.

Our study will contribute to the discussion of the benefits of using the Internet for political campaigning purposes by looking at candidates’ use of one of the most popular and most accessible services on the Internet for micro–blogging and online social networking, Twitter. Twitter has a user base of millions, reaching people right across the world and is used in traditional political campaigning (e.g., continuous and electoral campaigning) but also for non–traditional political activism (e.g., Iran in 2009, Tunis in 2010, Egypt in 2011). Furthermore, the influence of Twitter is felt beyond the its network since Twitter’s popularity caught the attention of mainstream media which now report regularly on politicians’ Twitter messages.

In our study, we will focus specifically on three major questions.

  1. To what extent did political candidates with different political backgrounds adopt micro–blogging as a campaigning tool in the European Parliament elections 2009?
  2. To what degree are these differences in political background related to the activity (micro–blogging and social networking) on the micro–blogging platform?
  3. To what extent is micro–blogging effective in gaining votes?

 

++++++++++

Hypotheses

The use of Web strategies

Establishment

Parties that were founded a long time ago as well as governing parties are viewed as being part of the establishment. These established parties, at least, appeal to a significant fraction of the electorate, enough to secure some basic and continued presence in Parliament. In general, leading politicians from these established parties receive a fair amount of media attention in the traditional media (television, newspaper and radio) (cf., Scholten and Ruigrok, 2006). Because of the limited space available in traditional media (time and square inches), the overrepresentation of established parties leaves new and small parties at a disadvantage. Since the Web solves this scarcity by providing virtually limitless space, new or less successful parties can use the Web as additional and alternative channels to increase visibility. Restrictions in terms of finance, time and space imposed by third parties (i.e., publishers and broadcasters) apply less in the Internet age.

Web 2.0 is very easy to use, mobile and the costs are very low, providing disadvantaged parties and candidates (e.g., few financial resources, little knowledge and small work force) increased opportunities to create more online visibility. New and smaller parties consist of politicians whose daily activities not only involve politics but also those of their regular jobs, which they still have to keep. Therefore, the benefit of the Web for smaller, less professionally organized parties and candidates might prove to be greater than for the more well–established parties and candidates. As such, the Web in general and micro–blogging specifically is believed to have the potential to overturn existing differences in general party and candidate visibility.

However, prior research does not show that less established parties benefit from the utilization of new media. According to Gibson and Ward (2009), larger parties in Parliament offer more sophisticated online campaigns than other parties. Exceptions are Green parties and far right parties that exploit new technology extensively. Chen and Smith (2010) also found that opposition parties use the Internet more often than incumbent parties. Therefore, contrary to the innovation hypothesis, the normalization hypothesis (Margolis, et al., 1999) states that the power distribution online is merely a replication of the offline power distribution. Larger parties might still have the upper hand because they have the advantage of a strategic department dealing continuously with publicity issues, and professional politicians that deal with politics as a daily business, as well as having greater experience.

This implies that established parties use new media technology more effectively than new, fringe parties do. To test whether the innovation hypothesis holds for Twitter, the following three–part hypothesis is formulated:

Hypothesis 1) Candidates from parties that do not belong to the establishment

  1. adopt micro–blogging more;
  2. micro–blog more frequently; and,
  3. perform more social networking

than candidates from parties that belong to the establishment.

Ideology

Ideology refers to ideas about how society should be organized, what societal goals should be achieved, and how to accomplish this. Keman (2007) argues that, in addition to the left versus right positioning of parties, another dimension is relevant: progressive versus conservative. Van Kersbergen and Krouwel (2008) identified both dimensions within the Dutch political system. Regarding left–wing versus right–wing ideology, it can also be argued that candidates from left–wing political parties are those candidates who are also the most communicative ones. Left–wing parties are more focused on a cohesive and supportive society, caring for weaker and disadvantaged groups in society. These characteristics might explain why left–wing parties use micro–blogging more extensively to connect to and engage with the electorate. However, an opposing argument is that more left–wing political parties promote collective action, whereas more right–wing thought propagates individual freedom. Applying this argument to electoral campaigning, one would expect that candidates from left–wing parties would conduct the campaign as a party collective and not individually. Right–wing parties on the other hand, would allow their candidates to design their campaign more freely and in an individualized way. Allowing candidates to use Twitter provides candidates with more freedom, less party control over how candidates perform in their campaign.

Furthermore, Gibson and Römmele (2001) suggest that right–wing parties are more willing to use a more business–like, professional campaigning approach, an approach the more left–wing parties dislike. Because it is not yet apparent whether candidates from left– or right–wing parties are more likely to adopt and use micro–blogging we pose the following research question instead of a hypothesis:

RQ 1: To what degree do left–wing and right–wing candidates use Twitter?

Regarding the progressive versus conservative ideological dimension, conservative thought promotes stability and continuity. In contrast, progressiveness promotes change and reform. Therefore, progressive parties and politicians are expected to adopt innovations more quickly than their conservative counterparts do. The hypothesis therefore is as follows:

Hypothesis 2) Candidates that are more progressive will adopt and use Twitter more than conservative candidates do.

Past electoral success

Gibson and Römmele (2001) suggest that a major party event such as the change of a party leader (internal shock) or a massive loss in elections (external shock) could affect a party’s campaign strategy. To try to overcome this shock in the upcoming elections, a new campaign strategy might be deployed, for instance, a new Web campaign strategy by utilizing new media technology such as Twitter. This then would increase the adoption of, for instance, micro–blogging and/or online social networking. The hypotheses therefore are:

Hypothesis 3) The smaller the electoral success in the past, the more it is likely candidates adopt and use Twitter.

Hypothesis 4) The more seats a party has lost in the last elections, the more likely it is that its candidates adopt and use Twitter.

Candidate characteristics

Not only can parties utilize the Web for increased visibility and better connectedness with voters, candidates can do this too. Particularly candidates who are ranked lower on the party list may benefit from using the Web for their personal campaign. In the Dutch electoral system, political parties prioritize their candidates from high to low. Normally candidates are elected for Parliament according to the party’s priority. However, voters may cast a preferential vote for a specific candidate. If this candidate receives enough preferential votes, he or she will be elected to Parliament, even though other candidates were given a higher priority by the party. Two major factors, besides the party program, that increase the likelihood of a candidate being elected are (1) the total number of votes the party receives, and (2) the number of preferential votes a candidate receives. If Web 2.0 is designed to be bottom–up, facilitating user generated content and creating a more level–playing field for all parties and candidates, it should be particularly beneficial to those who lack visibility and are the least likely to be elected. As such, personalized campaigning can be aimed at generating more preferential votes. The hypothesis then is as follows:

Hypothesis 5) The less priority the party has given a candidate the more likely it is that the candidates adopts and uses Twitter.

Research shows that women are more likely to use social network sites (Hargittai, 2007), but also report they have fewer Internet skills and perform less capital enhancing activities (Hargittai and Hinnant, 2008). Research on mobile phone use suggests that women use new media technology more sociably than men do (Ran and Lo, 2006). Because these findings on the role of gender in its relation to adoption of micro–blogging are somewhat contradictory, the research question is as follows:

RQ 2: To what extent is gender related to adoption and use of micro–blogging?

Effects of micro–blogging on the election outcome

One of the most important goals to use a particular election campaign strategy is to increase the number of votes for the party. Prior research shows evidence of Web campaigning to increase the number of votes: Gibson and McAllister (2006) show that having a Web site for campaigning purposes leads to more votes. Williams and Gulati (2008), focusing on social network sites, show that a candidate’s number of supporters on Facebook is related to that candidate’s vote share at the state level. We expect that candidates who blog more actively are more able to connect to potential voters, who subsequently are more likely to follow the candidate. Then, if potential voters are more likely to follow candidates, it is likely that the number of followers will be reflected in the number of votes for this particular candidate. The hypotheses then are as follows:

Hypothesis 6) The more a candidate uses micro–blogging, the more votes he or she receives.

Hypothesis 7) The larger a candidate’s online social network, the more votes he or she receives.

Besides these hypothesized relations, we will test for alternative explanations based on the level of establishment of parties and candidate prioritization. For instance, older parties, having a lot of experience and having proven to be an established actor in national politics, and having had electoral success in past elections, might receive more votes in the current elections than younger, fringe parties with little past performance. Furthermore, in the Dutch electoral system, which focuses on parties, the parties can order their candidates according the party’s candidate priority. It determines which candidates will be elected when no specific voter preference is indicated. If voters cast enough so–called preferential votes for a specific candidate, this can override the party’s candidate prioritization, resulting in this candidate’s election instead of another candidate. Still, the vast majority casts its vote on the #1 on the party list.

 

++++++++++

Methods

Data

The names of all candidates of political parties participating in the elections for the European Parliament as well as the ranking of the candidate on the party list and the number of votes he or she received were obtained from the Electoral Council (www.kiesraad.nl). Subsequently, by using various online sources (i.e., search engines, personal Web pages, political party Web pages), candidates who used Twitter prior to the elections were identified. This resulted in a list of 36 candidates, representing 12 of the 17 parties involved in the EU elections (see Table 1) running for a seat in the European Parliament and using Twitter. Six candidates had deleted their account, one of whom replaced his account with a new one. Data were downloaded in October 2009 from twitter.com using Twitter’s Application Programming Interface (API). The data were downloaded for the period of 1 February 2009 to 13 October 2009. However, for these analyses only data up to 4 June 2009 (i.e., Election Day) will be used.

Measurements

Adoption of micro–blogging site. Adoption of Twitter for micro-–logging was measured using two indicators: (1) whether or not a candidate was in possession of a Twitter account; and, (2) the start of micro–blogging was measured by the number of days prior to Election Day the first tweet was posted.

Micro–blogging activity was measured using several indicators. Message activity was measured by counting the total number of messages a candidate posted in the period from the official campaign started (17 March 2009) and the Election Day (4 June 2009). Message increase was measured by the average daily increase of tweets in the campaign. A low increase indicated that micro–blogging is performed in a normal regular fashion, whereas a large increase indicates that micro–blogging is used specifically for the campaign. The consistency of micro–blogging was measured by the standard deviation of daily micro–blogging activity: a low standard deviation indicates that the candidate has developed a routine of daily micro–blogging whereas a high standard deviation indicated that the candidate is an irregular micro–blogger.

To what degree the candidate is communicating with others on Twitter was measured by counting the number of @–tweets (i.e., messages sent to a specific twitterer). To standardize the @tweets and the non-@–tweets the communication ratio was calculated: lower than one indicates there were more non–directed tweets than directed tweets; larger than one indicates there were more directed tweets than non–directed tweets.

Networking activities. We distinguish a number of network characteristics. The candidate’s network size consists of all first–degree people in the network, regardless of people who are following the candidate or people from the public the candidate is following. The number of followers is the network size of people following the candidate while the number of following is the network size of people followed by the candidate. Reciprocal following is the number of following relations that are mutual between candidate and citizen. The candidate’s follower’s network size is the second–degree network size of those that follow the candidate (cf., the friend of a friend). The candidate’s following network size is the second–degree network size of those following others. The average number of shared connections is the mean number of identical relations candidates have with other people (irrespective of it being follower or following). Because the distribution of network indicators is heavily skewed, a log transformation was used to normalize them.

Establishment. Five indicators were used to measure whether parties are well–established. The first indicator is the age of the party (in years), measured by subtracting the year the party started using the party name. The date of formation was determined by the first appearance of a party under that specific name, excluding the years when parties were known under a different name. The second measurement is by its past popularity in elections, measured by the number of votes in the last national elections (i.e., 2006). The third indicator of whether candidates are member of a well–established political party was measured by whether the party participated in the national government since the last election, was in opposition or was considered a fringe party. The final indicator of the degree of establishment of political parties is the number of days a party was part of the national government.

Ideology was measured using data from Van Kersbergen and Krouwel (2008) who classified parties on 36 statements in the context of the 2006 general elections. An exploratory factor analysis was performed on these data, indicating seven dimensions (Eigen value > 1). However, only two dimensions were clearly interpretable: the left–wing versus right–wing continuum and the progressive versus conservative continuum.

Electoral success. The number of votes is the total number of valid votes per candidate. External electoral shock was measured by the change in number of seats in parliament a party obtained in 2006 as compared to 2003.

Candidate characteristics. The prioritization of the candidates by the political parties themselves measures the likelihood of a candidate being elected. The higher the candidate is ranked (indicated by a lower number), the more likely the candidate is elected.

Measures of central tendencies and dispersion of all variables are reported in the Appendix.

Analyses

Normally this type of research would involve multivariate analysis. However, due to a small sample and the lack of variance of activity for those without a Twitter account, combined with the large number of independent variables, using multivariate analysis risks overfitting the data resulting in trivial findings (Babyak, 2004). Therefore in this case it was not recommended. Consequently, only bivariate analyses are conducted, in particular the comparisons of means as well as the calculations of correlations.

 

++++++++++

Results

Establishment

Table 1 shows how party characteristics indicative for belonging to the political establishment are related to the adoption of micro–blogging (i.e., having a Twitter account and starting blogging early). Regarding whether the candidates’ parties are represented in government is unrelated to having a Twitter account (χ2 = 1.129, p = .524). However, candidates from parties in opposition started micro-blogging earlier than others, although the relation is not very strong (F = 2.484, p = .085). Candidates from parties that participated in the 2004 EP elections were not more likely to have a Twitter account (χ2 = 2.265, p = .116), but started micro–blogging somewhat earlier than others (F = 3.746, p = .054). The adoption of Twitter (i.e., subscription date to Twitter) and starting early with actual Twitter messaging are unrelated to the party’s participation in the general elections of 2006 (χ2 = 1.052, p = .305; F = 1.775, p = .184). Candidates who had a Twitter account were more often members of older parties than those that did not adopt micro–blogging (t = 7.853, p = .045), also candidates from older parties adopted micro–blogging earlier (r = .144, p = .027). Given these findings, hypothesis 1a receives limited support: candidates from older parties and those from parties in opposition signed up more frequently for Twitter, whereas candidates from fringe parties did not sign up very often.

Regarding the actual blogging activities (see Table 2) we see that especially candidates from parties in opposition blog more than other candidates (r = .367, p < .05), whereas candidates from fringe parties blog less frequently than others do (r = -.298, p < .05). This is also reflected in the directly sent messages to specific people (opposition: r = .255, p < .10; fringe: r = .282, p < .05). Furthermore, candidates who participated in prior campaigns blog more frequently than others (2004 campaign: r = .298, p < .05; 2006 campaign: r = .270, p < .10). This is also reflected in the directly sent messages (2004: r = .282, p < .05; 2006; r = .265, p < .10). As such, these results thus far provide no support for hypothesis 1b: candidates from fringe parties and candidates from parties that did not participate in prior EU elections did not micro–blog as extensively as those from parties already in the Dutch Parliament. Although candidates from governing parties are less active than candidates from opposition parties, candidates from fringe parties are even less active bloggers.

As for the network characteristics, Table 2 shows that especially candidates from government parties in general have smaller networks (range: r = -.226 – r = -.369, p < .10) and those from opposition and fringe parties have larger networks (range: r = .222 – r = .314, p < .10). Candidates who did not participate in the 2004 EP elections follow more people than those that did participate. These findings show mixed support for hypothesis 1c.

Ideology

Regarding the two ideologies (left wing versus right–wing and conservative versus progressive), Table 1 show no differences between those candidates with a Twitter account and those without. However, more progressive candidates adopted micro–blogging earlier than conservative ones (r = .200, p = .004). Reviewing the relations between actual micro–blogging activities and social networking activities, we find that progressive candidates blog more frequently (r = .323, p < .05), have a larger group of followers (r = .308, p < .05) and are more frequently involved in reciprocal relations on Twitter (r = .244, p < .10). Hypothesis 2 receives some support but not across the board.

 

Table 1: Analyses of the relation between the micro–blogging adoption and party and candidate characteristics.
Note: Nadopters=36, Nnon–adopters=252, except for the tests on ideology: Nadopters=30, Nnon–adopters=145; a. Mean difference=Madopters minus Mnon–adopters
 adopters
(those in possession of an account)
test statisticdfp–valuestart of micro–bloggingFp–value
Establishment 
Party type χ2=1.2912.524 2.484.085
governing (%)13.8%   mean=5.5  
opposition (%)14.2%   mean=12.4  
fringe (%)9.4%   mean=4.7  
Participated in EP ’04     3.745.054
yes (%)14.8%χ2=2.4651.116mean=10.8  
no (%)8.5%   mean=4.3  
Participated in EP ’06     1.775.184
yes (%)13.8%χ2=1.0521.305mean=9.8  
no (%)9.4%   mean=5.0  
# votes ’0673760at=-.474286.636r=-.022 .356
Age of party in ’097.853at=-2.017286.045r=-.114 .027
Ideology 
Left–wing versus right–wing.168at=-.923173.357r=.000 1.000
Conservative versus progressive.217at=-1.257173.210r=.200 .004
External shock 
Change in # of seats ’03–’06-1.647at=3.84958.000r=-.140 .009
Individual characteristics 
Candidate’s rank on party list-7.397at=7.52756.000r=-.229 .000
Gender     2.188.140
men (%)11.3%χ2=.7311.393mean=6.7  
women (%)14.9%   mean=11.8  

 

Election outcome

The results in Table 1 show that whether or not having a Twitter account or having started micro–blogging early are unrelated to number of votes their party received in the last 2006 elections (M difference = 73760, p = .636; r = .022, p = .356). However, according to the findings presented in Table 2, the number of votes received in 2006 elections is negatively correlated to network characteristics: the fewer votes a party received in the past, the larger the network sizes (range: r = -.222 – r = .404, p < .10) and the more frequently the candidates follow a member of the public or reciprocates a relationship on Twitter (r = .352, p < .05). These findings show little support for hypothesis 3.

Regarding the external shock hypothesis, candidates who had a Twitter account and those that started micro–blogging earlier were running for parties that had lost seats in Parliament in the last general elections (M difference = -1.647, p < .000; r = -.140, p = .009). The number of seats lost in the last elections however, is unrelated to blogging activities but has a positive relation with the candidate’s following net size (r = .226, p < .10). As such, hypothesis 4 receives little support.

Individual characteristics

The priority a party has given a candidate is related to having an account: those candidates with a Twitter account and started micro–blogging early on were more prioritized (i.e., lower rank number) by the party (M difference = 7.397, p < .000 r = -.229, p < .000). Candidates who were higher prioritized by the party (i.e., a lower number) blogged more inconsistently (i.e., a larger standards deviation) throughout the campaign period (r = .363, p < .05). These higher prioritized candidates also showed significantly larger networks (range: r = -.220 - r = .388, p < .10). These finding offer no support for hypothesis 5, which predicted the disadvantaged candidates would utilize micro–blogging more actively.

Gender was unrelated to having a Twitter account (χ2 = .731, p = .393) as well as to the number of days ago the candidate started blogging (F = 2.188, p = .140). Gender was also unrelated to blogging activities and to networking activities (RQ 2).

 

Table 2: Correlates of party and candidate characteristics and micro–blogging activities and network characteristics.
Note: * p<.10, ** p<05, *** p<.01; N=36, except for correlations involving political ideologies: N=30; a. Natural log transformation; b. None of the fringe parties in 2009 participated in the European Parliament elections of 2004. This results in a perfect negative correlation between ‘fringe party’ and ‘participation in 2004 EP elections’. This in turn results in perfectly mirrored correlations of these variables with micro–blogging activities and network characteristics.
 Blog characteristicsNetwork characteristicsa
# number of tweetsSD daily tweetsDaily tweet increaseFrequency of @–tweetsRatio @tweet/non@tweetCandidate’s followers net sizeCandidate’s following net sizeAverage # shared connectionsFollower of candidateFollowing member of publicReciprocal following
Establishment 
party type 
governing-.130-.116.154-.013.046-.297**-.369**-.086-.250*-.226*-.313**
opposition.367**.215-.058.255*.140.142.115-.021.314**-.056.068
fringe-.298**-.137-.081-.282**-.205.121.222*.107-.122.281**.223*
# votes GE ’06-.059.041.176.081.161-.368**-.404***-.188-.222*-.279**-.352**
participated in EP ’04b.298**.137.081.282**.205-.121-.222*-.107.122-.281**-.223*
participated in EP ’06.270*.142.056.265*.213-.092-.178-.058.118-.213-.180
age of party.170.175.015.218.253*-.169-.213-.109-.001-.213-.170
External shock 
change in # seats 2003–2006-.054-.108-.019-.120-.161.198.226*.113.070.199.138
Political ideology 
left–wing versus right–wing-.076.174.042.022.184-.079-.055-.168-.223.014.002
conservative versus progressive.323**.186-.025.221.071.154.152-.075.308**.120.244
Candidate characteristics 
candidate’s rank on party list-.216-.363**-.271*-.233*-.158-.254*-.220*-.185-.132-.388*-.204
gender-.194-.053-.043-.078.025-.016-.047.029-.086.022-.058

 

The effects of micro–blogging on election outcome

The first column in Table 3 shows the correlations of blogging activities and networking activities with the number of votes candidates received in the European Parliament elections of 2009 for all candidates. The more frequently candidates tweeted, the more votes they received (r = .318, p < .01). The less consistently (i.e., larger standard deviation) candidates blogged, the more votes they received (r = .422, p < .01). Candidates who increased their blogging closer to Election Day also received more votes (r = .307, p < .01). Furthermore, the more messages were sent to a specific Twitter account (@tweets), the more votes they received. These findings suggest that using micro–blogging as campaign tool is effective in getting more votes, supporting hypothesis 6.

Notable is that the correlations between blogging activities and the number of votes received for the subsample of micro–blogging adopters (third column) are similar to the findings in the second column (all candidates). This is different for network characteristics: nearly all correlations between network characteristics and the number of votes are non–significant.

Regarding the networking activities based on all candidates, only the number of people following the candidate was related positively to the number of votes (range r = .312 – r = .426, p < .01). However, the correlations between networking activities and the number of votes, only for candidates having an account, shows merely one significant correlation: those that have a large following received more votes (r = .328, p < .10), a finding offering only limited support for hypothesis 7.

Although these findings seem to support our expectations, there might be alternative explanations for the relations we found between micro–blogging activities and Twitter network characteristics. Therefore we calculated correlations between party characteristics, past–electoral success, and candidate characteristics. All party characteristics show significant relations with the number of votes received in the 2009 EP elections. Candidates from parties represented in government received more votes, as did those that participated in the 2004 EP elections and the 2006 general elections. As for the two ideologies, the more right–wing the candidate’s ideology, the fewer votes these candidates receive. Regarding the conservative–progressive dimension there is no relation with the number of votes candidates received. The number of votes the party received in the general elections of 2006 correlated positively with the number of votes the candidate received in the 2009 EP elections. The rank number given by the party to the candidate in the 2009 EP election (indicating the level of priority) correlated negatively with the number of votes they received: the more priority the candidate has been given by the party, the more votes he or she receives. The existence of these substantial correlations suggests that they should be considered as alternative explanations, possibly even able to declare earlier discussed relations between micro–blogging and social networking as being spurious.

 

Table 3: Correlation analysis of cast votes, party and candidate characteristics and micro–blogging activities.
Note: * p<.10, ** p<05, *** p<.01; a. Natural log transformation.
 Votes per candidate
all candidates
(n=288)
subsample of micro–blogging candidates
(n=36)
Adoption 
Tweet start (# days prior to Election Day).318***-.068
Blog activity 
# of tweets.342***.310*
Standard deviation daily tweets.422***.511***
Daily tweet increase.307***.306*
Frequency of @tweets.320***.337**
Ratio @tweets/non@tweets.031.343**
Network characteristicsa 
# followers of the party or candidate.426***.328*
# following a member of the public.332***.009
Reciprocal following.312***-.129
Candidate’s followers net size.394***.115
Candidate’s following net size.385***.045
Average # shared connections (network overlap).354***.097
Establishment 
Party type 
Ruling parties.345*** 
Opposition parties.320*** 
Fringe parties-.632*** 
Participated in EP elections 2004.537*** 
Participated in general elections 2006.593*** 
# votes general elections 2006.500*** 
Age of the party in 2009.412*** 
Ideology 
Left–wing versus right–wing-.132* 
Conservative versus progressive-.007 
External shock 
Change in number of seats 2003–2006-.157*** 
Candidate characteristics 
Candidate’s rank on party list-.324*** 
Gender-.203*** 

 

 

++++++++++

Discussion and conclusions

This study focused on (1) acquiring a Twitter account and (2) activities of micro–blogging and social networking in political campaigning. Subsequently (3) we explored whether there were indications of the effectiveness of micro–blogging and social networking activities in gaining more votes. Micro–blogging as a campaigning tool was adopted only by a minority of approximately 13 percent of the EP candidates in 2009. Regarding candidates’ activities on the micro–blogging site Twitter, the results vary quite extensively. Overall, candidates from opposition parties started micro–blogging significantly earlier, sending more messages and securing larger follower networks. Candidates from fringe parties demonstrated lower blogging activity, but had high networking activity, which was often reciprocated. However, fringe parties and lower ranked candidates were not able to utilize micro–blogging for their own benefit. This was surprising, especially, because the use is free and acquiring Twitter skills is relatively easy. This suggests that a lack of previous campaign experience limited the adoption of new campaigning tools, such as those provided on the Web. Alternatively, limited resources (e.g., finance, time and knowledge) of fringe parties and candidates might explain this. Candidates from parties with a history of campaigning in prior elections were more active in using micro–blogging. At the same time, candidates from unsuccessful parties (i.e., fewer votes in prior elections) used micro–blogging more extensively. As such, it seemed that micro–blogging was associated with being in a disadvantaged position, in particular opposition parties, but excluding fringe parties. These findings point towards innovative use of Web campaigning by disadvantaged parties and candidates. The parties in power (i.e., governing parties) consolidated their campaign efforts: these candidates demonstrated very little adoption of and activity on Twitter.

What are the benefits of micro-blogging has for politics in general? How would a specific politician regard micro–blogging as effective? In this study, we looked at a central outcome variable: the number of votes a candidate received. The results clearly show positive relations with micro–blogging use and the number of votes measured over the entire population of candidates. However, because the adoption rate is quite small, these activity indicators measured over the entire population reflect the adoption of Twitter and not actual activity. Looking at the subsample of Twitter adopters message activities and the number of votes show positive relations, but relations between network characteristics and the number of votes are virtually absent. As such, the Twitter network size seems to be a limited indicator for voting outcome. Apparently, the composition of these online networks differs from groups of people that vote for specific candidates, suggesting the degree of homophily (cf., McPherson, et al., 2001) is low. This raises the question whether Norris’ (2003) claim that online social networking means ‘preaching to the converted’ instead of ‘preaching to the disbelievers’ is still valid. If these networks indeed differ, creating online social networks could be worthwhile to expand existing off–line networks. However, these findings also suggest that online networks do not lead to more votes.

Although having established a relation between micro–blogging and the number of votes, one must take into account alternative explanations indicated by prior electoral success and the candidate’s priority: these also show significant relations with the number of votes these candidates received. This suggests that the existing relations could turn out to be spurious. To test this, a larger sample would be needed, enabling multivariate analyses. This study then should be replicated in the future, preferably using data sets from different countries.

If the answer to this study’s main question “Is the voter only a tweet away” is a definite “yes”, i.e., there is indeed a real effect of micro–blogging activity on the number of votes candidates receive (as is suggested by the findings in this study), the question arises how this should be interpreted and explained. Is it because followers of the candidates are better informed than those without Twitter, that candidates using Twitter were able to convince voters to change their vote? We assume that conveying complex messages on complicated political issues by using merely 140 characters is quite difficult, unless one uses many messages or (shortened) hyperlinks to direct followers to more informative Web pages. Alternatively, are people voting for a particular candidate because the candidate managed to present him or herself as more likeable or more approachable? Apparently people appreciate becoming friends with politicians on social network sites, even though this constitutes a weak tie for bridging purposes as opposed to a strong tie for bonding purposes (Ellison, et al., 2007). Thus far, the few existing studies show limited effects. Utz [2] demonstrated that visitors to a politician’s SNS page, noticing that a politician indicated reacted to online comments, tended to have a more favorable attitude towards a given politician. Baumgartner and Morris (2010) on the other hand, demonstrated that the potential for SNS to increase political engagement was limited at best: young people used SNS to share pre–existing ideas instead of acquiring new ones. Furthermore, the ability of youngsters to learn about politics and candidates from SNS was limited. Williams and Gulati (2008) also found a relation between a SNS (Facebook) and election outcome: the higher the percentage of supporters for a particular candidate on Facebook per U.S. state, the larger the vote share of that candidate in that U.S. state.

Even though effects seem to be limited, questions about knowledge transfer interactivity, personalization and mobilization by new media become more relevant in an age of decreasing political involvement. Politicians feel the need to find new and more ways to reach citizens in general and potential voters in particular. Still, old channels having a wide and diverse audience (interpersonal, television, radio and newspaper) are still being used. However, if the popularity of social media use is inherited by future generations and people are increasingly able to avoid information that conflicts with their preexisting knowledge, democracy is at risk. End of article

 

About the authors

Maurice Vergeer is associate professor at the Department of Communication at the Radboud University, Nijmegen, the Netherlands, and visiting professor of the WCU Webometrics Institute at Yeungnam University, South Korea. His research interests focus on online communication and networks and online participation, in particular in political communication and social capital theory.
Web: www.mauricevergeer.nl
E–mail: m [dot] vergeer [at] maw [dot] ru [dot] nl

Liesbeth Hermans is associate professor at the Department of Communication at the Radboud University in Nijmegen. Her field of research is journalism studies and political communication. Her interests in journalism include the influence of digitalization on the news process, role perceptions of journalists, changing journalistic routines and conditions because of new technology, and the changing relation with the audience.
E–mail: l [dot] hermans [at] maw [dot] ru [dot] nl

Steven Sams is a doctoral candidate at the School of Information Systems, Computing, and Mathematics at Brunel University and a research fellow of WCU Webometrics Institute at Yeungnam University. Currently he focuses on data collection strategies for the analysis of political communication in South Korea.
E–mail: steven [dot] sams [at] gmail [dot] com

 

Acknowledgements

This research was supported by WCU (World Class University) program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (No. 515–82–06574).

 

Notes

1. Schweitzer, 2008, p. 458.

2. Utz, 2009, p. 235.

 

References

Jody C. Baumgartner and Jonathan S. Morris, 2010. “MyFaceTube politics social networking Web sites and political engagement of young adults,” Social Science Computer Review, volume 28, number 1, pp. 24–44.http://dx.doi.org/10.1177/0894439309334325

Michael A. Babyak, 2004. “What you see may not be what you get: A brief, nontechnical introduction to overfitting in regression–type models,” Psychosomatic Medicine, volume 66, number 3, pp. 411–421.http://dx.doi.org/10.1097/01.psy.0000127692.23278.a9

Benjamin R. Barber, Kevin Mattson and John Peterson, 1997. The state of ‘electronically enhanced democracy’: A survey of the Internet. New Brunswick, N.J.: Walt Whitman Center for the Culture and Politics of Democracy, Department of Political Science, Rutgers University.

Gian Vittorio Caprara, Claudio Barbaranelli and Philip G. Zimbardo, 1999. “Personality profiles and political parties,” Political Psychology, volume 20, number 1, pp. 175–197.http://dx.doi.org/10.1111/0162-895X.00141

Peter Chen and Peter Jay Smith, 2010. “Adoption and use of digital media in election campaigns: Austria, Canada, New Zealand,” Public Communication Review, volume 1, number 1, at http://epress.lib.uts.edu.au/ojs/index.php/pcr/article/view/1249, accessed 11 July 2011.

Graham Cormode and Balachander Krishnamurthy, 2008. “Key differences between Web 1.0 and Web 2.0,” First Monday, volume 13, number 6, at http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/2125/1972, accessed 28 February 2009.

Electoral Council, 2009. “Proces verbaal EP verkiezingen 2009 [Official record EP elections 2009],” at http://www.kiesraad.nl/nl/Overige_Content/Bestanden/pdf_thema/Proces-verbaal_EP_verkiezingen_2009.pdf, accessed 7 January 2010.

Homero Gil De Zúñiga, Eulàlia Puig–I–Abril, and Hernamdo Rojas, 2009. “Weblogs, traditional sources online and political participation: An assessment of how the Internet is changing the political environment,” New Media & Society, volume 11, number 4, pp. 553–574.http://dx.doi.org/10.1177/1461444809102960

Nicole Ellison, Charles Steinfield and Cliff Lampe, 2007. “The benefits of Facebook ‘friends’: Social capital and college students’ use of online social network sites,” Journal of Computer–Mediated Communication, volume 12, number 4, at http://jcmc.indiana.edu/vol12/issue4/ellison.html, accessed 11 July 2011.

European Commission, 2009. EUROBAROMETER 71: Public opinion in the European Union. Brussels: European Commission, at http://ec.europa.eu/public_opinion/archives/eb/eb71/eb71_std_part1.pdf, accessed 11 July 2011.

European Parliament, 2009. “Turnout at the European elections (1979–2009),” at http://www.europarl.europa.eu/parliament/archive/staticDisplay.do;jsessionid=5B3BDBB835DEB4A5658D405AB8E364CC.node1?language=EN&id=211, accessed 17 January 2010.

Richard S. Flickinger and Donley T. Studlar, 2007. “One Europe, many electorates? Models of turnout in European Parliament elections after 2004,” Comparative Political Studies, volume 40, number 4, pp. 383–404.http://dx.doi.org/10.1177/0010414006288970

Kirsten A. Foot, Steven M. Schneider, Randolph Kluver, Michael Xenos and Nicholas W. Jankowski, 2007. “Comparing Web production practices across electoral Web spheres,” In: Randolph Kluver, Nicholas W. Jankowski, Kirsten A, Foot and Steven M. Schneider (editors). The Internet and national elections: A comparative study of Web campaigning. London: Routledge, pp. 243–260.

Rachel Gibson and Stephen Ward, 2009. “Parties in the digital age: A review article,” Representation, volume 45, number 1, pp. 87–100.http://dx.doi.org/10.1080/00344890802710888

Rachel Gibson and Ian McAllister, 2006. “Does cyber–campaigning win votes? Online communication in the 2004 Australian election,” Journal of Elections, Public Opinion & Parties, volume 16, number 3, pp. 243–263.http://dx.doi.org/10.1080/13689880600950527

Rachel Gibson and Andrea Römmele, 2001. “Changing campaign communications: A party–centered theory of professionalized campaigning,” Harvard International Journal of Press/Politics, volume 6, number 4, pp. 31–43.http://dx.doi.org/10.1177/108118001129172323

Vassia Gueorguieva, 2008. “Voters, MySpace, and YouTube: The impact of alternative communication channels on the 2006 election cycle and beyond,” Social Science Computer Review, volume 26, number 3, pp. 288–300.http://dx.doi.org/10.1177/0894439307305636

Girish J. Gulati and Christine B. Williams, 2010. “Congressional candidates’ use of YouTube in 2008: Its frequency and rationale” Journal of Information Technology and Politics, volume 7, numbers 2–3, pp. 93–109.

Eszter Hargittai, 2007. “Whose space? Differences among users and non–users of social network sites,” Journal of Computer–Mediated Communication, volume 13, number 1, at http://jcmc.indiana.edu/vol13/issue1/hargittai.html, accessed 11 July 2011.

Eszter Hargittai and Amanda Hinnant, 2008. “Digital inequality: Differences in young adults’ use of the Internet,” Communication Research, volume 35, number 5, pp. 602–621.http://dx.doi.org/10.1177/0093650208321782

Hans Keman, 2007. “Experts and manifestos: Different sources — Same results for comparative research?” Electoral Studies, volume 26, number 1, pp. 76–89.http://dx.doi.org/10.1016/j.electstud.2006.04.004

Randolph Kluver, Nicholas W. Jankowski, Kirsten A, Foot and Steven M. Schneider (editors), 2007. The Internet and national elections: A comparative study of Web campaigning. London: Routledge.

Anders Olof Larsson, 2011. “‘Extended infomercials’ or ‘Politics 2.0’? A study of Swedish political party Web sites before, during and after the 2010 election,” First Monday, volume 16, number 4, at http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/3456/2858, accessed 13 June 2011.

Michael Margolis, David Resnick and Joel D. Wolfe, 1999. “Party competition on the Internet in the United States and Britain,” Harvard International Journal of Press/Politics, volume 4, number 4, pp. 24–47.http://dx.doi.org/10.1177/1081180X9900400403

Miller McPherson, Lynn Smith–Lovin and James M. Cook, 2001. “Birds of a feather: Homophily in social networks,” Annual Review of Sociology, volume 27, pp. 415–444.http://dx.doi.org/10.1146/annurev.soc.27.1.415

Pippa Norris, 2003. “Preaching to the converted? Pluralism, participation and party Web sites,” Party Politics, volume 9, number 1, pp. 21–45.http://dx.doi.org/10.1177/135406880391003

Pippa Norris, 2001. Digital divide? Civic engagement, information poverty and the Internet worldwide. Cambridge: Cambridge University Press.

Tim O’Reilly, 2005. “What is Web 2.0? Design patterns and business models for the next generation of software” (30 September), at http://oreilly.com/web2/archive/what-is-web-20.html, accessed 23 March 2007.

Wei Ran and Ven–Whei Lo, 2006. “Staying connected while on the move: Cell phone use and social connectedness,” New Media & Society, volume 8, number 1, pp. 53–72.http://dx.doi.org/10.1177/1461444806059870

Otto Scholten and Nel Ruigrok, 2006. Politiek en politici in het nieuws in vijf landelijke dagbladen. Amsterdam: Stichting het Persinstituut de Nederlandse Nieuwsmonitor, UvA.

Eva Johanna Schweitzer, 2008. “Innovation or normalization in e–campaigning? A longitudinal content and structural analysis of German party Web sites in the 2002 and 2005 national elections,” European Journal of Communication, volume 23, number 4, pp. 449–470.http://dx.doi.org/10.1177/0267323108096994

John C. Tedesco, 2004. “Changing the channel: Use of the Internet for communicating about politics,” In: Lynda Lee Kaid (editor). Handbook of political communication research. Mahwah, N.J.: Lawrence Erlbaum Associates, pp. 507–532.

Sonja Utz, 2009. “The (potential) benefits of campaigning via social network sites,” Journal of Computer–Mediated Communication, volume 14, number 2, pp. 221–243.http://dx.doi.org/10.1111/j.1083-6101.2009.01438.x

Kees Van Kersbergen and Andre Krouwel, 2008. “A double–edged sword! The Dutch centre–right and the ‘foreigners issue’,” Journal of European Public Policy, volume 15, number 3, pp. 398–414.http://dx.doi.org/10.1080/13501760701847655

Renée van Os, Nicholas W. Jankowski and Maurice Vergeer, 2007. “Political communication about Europe on the Internet during the 2004 European Parliament election campaign in nine EU member states,” European Societies, volume 9, number 5, pp. 755–775.http://dx.doi.org/10.1080/14616690701412871

Maurice Vergeer, Liesbeth Hermans and Carlos Cunha, forthcoming. “Political parties, candidates, and Web campaigning in the 2009 European Parliament elections: A decade of cross–national comparative Web feature analysis,” New Media & Society.

Christine Williams and Girish J. Gulati, 2008. “What is a social network worth? Facebook and vote share in the 2008 Presidential primaries,” paper presented at the 2008 annual meeting of the American Political Science Association, Boston, Mass. (28–31 August), at http://blogsandwikis.bentley.edu/politechmedia/wp-content/uploads/2008/10/apr-sept-1.pdf, accessed 22 September 2010.

 

Appendix

 

Descriptive measures for the (sub)population of candidates in the European Parliament elections of 2009.
 All candidates; (N=288) Candidates possessing an account; (N=36)
 MinimumMaximumMeanSD MinimumMaximumMeanSD
 
Blogging activity 
Candidates with an account
(no=0, yes=1)
01.00.10.3 1.01.01.00
Number of days active account0151.08.427.6 0151.067.147.1
# of messages01378.031.2139.4 01378.0249.3321.7
Standard deviation daily tweets017.40.41.6 017.43.33.5
Daily tweets increase-0.00.40.00.0 -0.00.40.00.1
Frequency of directly addressed messages07.80.20.8 07.81.32.0
Communication ratio03.51.00.3 03.50.80.8
 
Network characteristics 
# followers of the party or candidate02723.041.8222.4 02723.0334.8552.0
# following a member of the public01296.011.784.2 01296.093.3224.3
# of reciprocal following01282.016.6103.1 01282.0132.9267.0
Candidate’s followers network size (2nd degree) (*1000)0187346.31103.111285.3 0187346.38825.231212.7
Candidate’s following network size (2nd degree) (*1000)013620.5142.2991.0 013620.51137.62624.7
Average # shared connections08.50.41.4 08.53.52.5
 
Establishment 
Candidates from 
ruling parties (no=0, yes=1)01.00.20.4 01.00.20.4
opposition parties (no=0, yes=1)01.00.50.5 01.00.50.5
fringe parties (no=0, yes=1)01.00.30.5 01.00.30.4
Participated in European Parliament elections 2004 (no=0, yes=1)01.00.60.5 01.00.80.4
Participated in general elections 2006 (no=0, yes=1)01.00.70.5 01.00.80.4
Age of the party in 2009063.023.522.0 063.030.323.2
# votes (*1000) in general elections 200602608.6758.2871.8 02608.6822.7899.2
 
External shock 
Change in # seats-9.09.0-1.33.3 -9.02.0-2.73.1
 
Ideology 
Left–wing versus right–wing-1.31.4-0.00.9 -1.31.40.11.0
Conservative versus progressive-1.41.10.10.9 -1.41.10.30.9
 
Individual characteristics 
Candidate’s rank on party list1.030.011.27.7 1.020.04.85.1
Gender (female=0, male=1)01.00.70.5 01.00.60.5
 
Voting outcome 
# votes (*1000) in EP elections 20090579.815.865.1 0579.880.2149.0

 

 


Editorial history

Received 25 April 2011; revised 11 June 2011; accepted 4 July 2011.


Copyright © 2011, First Monday.
Copyright © 2011, Maurice Vergeer, Liesbeth Hermans, and Steven Sams.

Is the voter only a tweet away? Micro–blogging during the 2009 European Parliament election campaign in the Netherlands
by Maurice Vergeer, Liesbeth Hermans, and Steven Sams.
First Monday, Volume 16, Number 8 - 1 August 2011
http://firstmonday.org/ojs/index.php/fm/article/view/3540/3026





A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2016.