Social Media: The Fifth Estate

Recent technological changes have lead to the rapid development of communication networks through social media, with political implications that are only beginning to be understood.  Approximately 80% of American adults use the Internet and 66% of those use social network sites (SNS), such as Facebook, LinkedIn, Twitter, Google+, etc. About the 75% of social network users in U.S. tend to share posts related to politics and political news with their contacts, and most importantly the 25% of SNS users say they have become more active in politics, and the 16% of SNS users say they have changed their political views after discussing it or reading posts about it on the sites (Pew Research Centre, 2016).

The higher level of exposure to political information is often mediate by SNS. Indeed, while individuals are nowadays easily overexposed to information and they have virtually access to any source of information, the new technologies also allow them to restrict their exposure only to like-minded sources, this way creating a sort of ``filtered environment''.  In turn, this has created a major paradox for the standard model of political competition in environments where strategic information is important. Specifically, this paradox revolves around the fact that while new technologies have massively increased the availability of information to voters (via direct news, social media exchanges or targeted political advertising) they also seem to have facilitated the development of segregated information environments – so-called echo chambers. These echo chambers are a potential problem because they prevent voters from acquiring as much information as they feasibly could. As result, valuable information may be wasted or disregarded.

In 1944, Lazarsfeld’s seminal paper challenged the conventional idea of a direct influence of the mass media on the general public by introducing the “two-step hypothesis of communication flows”: viewpoints, ideas and political programs flow from mass media to opinion leaders and from them to voters. Thus, while some studies focusing on the consumption of ‘top-down’ news (eg: print newspapers, TV, online news websites) found evidence of higher information segregation (Gentzkow and Shapiro, 2011), more recent empirical studies have focused on word-of-mouth communication (voter-to-voter communication) and have provided evidence that the decision-making processes occur within voter personal networks (McClurg, 2003; Plutzer and Zipp, 1996). They find that interactions within particular types of networks, such as those based on intimate social relations, are very likely to foster a greater collective interest in politics. Bakshy, Messing, and Adamic (2015) investigate how U.S. Facebook users interact with socially shared news (sharing activity) and examine the extent to which heterogeneous Facebook friends could expose individuals in their network to cross-cutting content (material that is not in line with their political biases). They find that political post shared by individuals play a stronger role than the Facebook recommendation algorithm in limiting exposure to cross-cutting content. In other words, social contact matters playing the role of information ‘filter’ for potential voters.

Theoretical approaches to this problem can be divided in two arguments.  The first argument focuses on the ``top-down’’ approach and analyses the effects of informative advertising on political competition. There is not yet a consensus in this literature: while Cox and McCubbins (1986) and Galeotti and Mattozzi (2011) find that parties tend to target their transfers on their own “core supporters”, who are easy to ‘buy’, Lindbeck and Weibull (1987) and Schulz (2003) predict that parties are more likely to spend resources for targeting “swing voters”, i.e. the voters who can switch their vote. 

The second argument is based on ``bottom-up’’ models; it focuses on the effect of personal influence (information transmission among voters) on voting decisions. When voters have similar preferences, voters’ messages directly affect their voting decision (Austen-Smith, 1990), but when voters have conflicting preferences the information gathered from other voters is used as a double-check mechanism: voters will switch their vote if and only if their private information is confirmed by the message received (Doraszelski, Gerardi, and Squintani, 2003).

Clearly, the new information environment suggests that new approaches are needed to study political competition amongst parties. Specifically, these approaches arguably need to combine the classic top-down model, in which parties use advertising to persuade voters, with a bottom-up model that takes into account the new role of voter-to-voter communication in shaping public opinion and influencing voting decisions. The potential theoretical basis of this phenomenon has been investigated in Giovanniello (2017), which embeds (i) a model of campaign competition, where parties compete in campaign advertising to convince voters to cast their vote for them (top-down approach), and (ii) a model of personal influence where voters can strategically decide which information (if any) they want to share with their network to manipulate the political outcome (bottom-up approach).

As a practical example of voters communication consider the case of left-right or Democrat-Republican competition with two candidates, say Obama and Romney. There are two voters who are matched as sender and receiver, and the receiver’s vote is pivotal, i.e., she will decide the outcome of the election. Assume both voters want to elect a moderate candidate, but they will always prefer to elect a moderate candidate who is affiliated to their favored party.

Following this, suppose that the sender knows that Romney is a moderate but she has no information about the type of Obama. Then, a Democrat sender matched with Democrat receiver will always truthfully reveal her information about both candidates. In fact a Democrat receiver will vote for the moderate Romney if and only if she has no information at all about Obama’s type, but will vote for Obama otherwise. However, if a Democrat sender is matched with a Republican receiver and she truthfully reveals her information, then the Republican receiver will always vote for Romney even if she knows that Obama is a moderate. By “lying” to the Republican receiver, and not revealing Romney’s type, a Democrat sender may therefore increase the probability that the receiver will vote for a moderate Democratic candidate (Obama). Hence, Democrat-biased voters truthfully convey their information to other Democrat-biased voters, but will not convey any information otherwise. In short, Democrats only talk (truthfully) to Democrats and similarly, Republicans only talk (truthfully) to other Republicans.

Thus, when voters can strategically communicate with each others echo chambers emerge endogenously: in equilibrium, voters only share valuable information with like-minded peers who are biased toward their own favored party. In other words, two voters who are ideologically close but biased toward different parties, for example a moderate left voter and a moderate right voter, will never share truthful information; while two voters who are ideologically distant but biased toward the same party, for example a left independent voter and a left moderate partisan, will truthfully share their private information. In turn, this prevents the acquisition of valuable information among voters who are biased toward the other party.

The rise of the echo chambers, via the restricted diffusion of information, highlights how information flows can impinge on electoral competitions. In turn, this makes it crucial for parties to design their advertising strategies in a way that exploits these echo chambers so as to increase their probability of winning elections. . This result drives the second finding of Giovanniello (2017): if parties have access to targeting technologies, they may prefer to target voters that are ideologically biased toward their opponent, rather than their own supporters, and exploit the echo chambers effect.

The second result suggests that both the richness of the network and the degree of homophily play an important role in the advertising strategies of the parties. Specifically, whenever either the richness of the network, or the degree of homophily within the network or both, are low, the parties are likely to tailor their advertising campaign to voters ideologically biased toward their opponent - rather than targeting the closer ideological group of voters, as the literature with truthful communication suggests.

The crucial contribution of Giovanniello (2017) lies in its ability to disentangle the mechanisms throughout communication flows, within contemporary social media networks, shapes the electoral competition. The two results described above suggest that political interaction in social networks cannot be ignored, neither can be ignored the role played by the network structure. Understanding how social network shape voting decisions has become crucially important also for the mobilization of the so-called independent voters, i.e., voters who are likely to base their voting decisions only on candidates’ personal characteristics, that are decisive for election outcomes. So far the role played by network structure in the political arena has been comparatively little examined.

References

1. Austen-Smith D., (1987), ”Interest groups, campaign contributions and probabilistic voting”, Public Choice 54(2): 123-139
2. E. Bakshy, S. Messing, and L.A. Adamic (2015).“Exposure to ideologically diverse news and opinion on Facebook” Science 348: 1130–1132
3. Cox G. W., and McCubbins M. D., (1986), ”Electoral politics as a redistributive game”, The Journal of Politics 48(2), 370-389
4. Doraszelski U., Gerardi D., and Squintani F., (2003), ”Communication and voting with double-sided information”, Contributions in Theoretical Economics 3(1).
5. Galeotti A., and Mattozzi A. (2011), ”Personal Influence : Social Context and Political Competition”, American Economic Journal: Microeconomics.  
6. Giovanniello M.A. (2017), “Echo Chambers: Voter-to-Voter Communication and Political Competition”, mimeo
7. Gentzkow M., Shapiro J. M., (2011), “Ideological segregation on-line and off-line”, 
The Quarterly Journal of Economics 126, 1799-1839.
8. Lazarsfeld P.F., Berelson B., and Gaudet H., (1944), ”The people’s choice: how the voter makes up his mind in a presidential campaign”, Columbia University Press
9. Lindbeck A., and Weibull J., (1987), ”Balanced-budget redistribution as the out- come of political competition”, Public Choice 52(3): 273-297
10. McClurg Scott D., (2003), ”Social Networks and Political Participation: The Role of Social Interaction in Explaining Political Participation”, Publications. Paper 6
11. Plutzer E., and Zipp J. F., (1996), ”Identity politics, partisanship, and voting for women candidates”, Public Opinion Quarterly 60(1):30-57
12. Rainie L.,and Smith A. (2012), ”Social networking sites and politics”, Pew Internet Project
13. Schultz C., (2003), ”Strategic Campaigns and Redistributive Politics”, mimeo

​