During the 2010 midterm elections, a group of researchers found that Twitter posts could pretty accurately predict who win congressional elections. Although the social networking site has been around since 2006, political scientists have done little to tie Twitter to political behavior.

In any election, the two major party candidates are likely to mentioned on Twitter. However, this article finds that the candidate who is mentioned more on the social networking service is likely to receive more votes. The effect is equal to 1,035 votes for every percentage increase in Twitter mentions over the other candidate.

While it’s nice to see this quantified, the findings are not surprising. Most congressional elections are not competitive. In 2010, 86% of officeholders in Congress were re-elected to office. In terms of mentions on Twitter, I would expect incumbents to receive far more tweets than challengers.

This article, “More Tweets, More Votes: Social media as a quantitative indicator of political behavior” was presented at the American Sociological Association’s Annual Meeting in New York, NY on August 10-13, 2013. (Link)

As a researcher, I have co-authored a paper that investigates political Facebook use amongst younger age cohorts. Stephen Wolfram, author of Mathematica, released a series of different statistics on Facebook usage using a sample that includes over a million people.

A small fraction of this data examines the political use of Facebook:

Figure 1. Age distribution of Facebook users compared to the US Census.

Figure 2. The number of friends based on age.

Figure 3. Age distributions of users based on age.

Figure 4. The popularity of topics discussed in status updates.

Figure 5. Status update word cloud: politics.

Figure 6. The discussion of politics based on age and gender.

These graphs, taken together, seem to be telling an interesting story. Figure 1 shows that Facebook seems to be in strong popularity amongst younger users. This fact is not immediately surprising. In addition, Figure 2 shows that younger users have more friends.

Having more friends is not necessarily a good thing. On average, Facebook users in their 20s have 400 friends. Here, it’s appropriate to cite the Dunbar number, suggesting humans can only maintain 150 meaningful relationships at one time. Younger Facebook users may be collecting friends, while perhaps older users spend more time fostering the fewer relationships they have on the social networking site.

Figure 3 shows the distribution of ages for users based on their own age cohort. Most users have a strong base of friends that are their own age, the only exception being users over the age of 70. Another surprising finding of Figure 3 suggests that users between the ages of 50 and 70 have a substantial proportion younger friends. Thus, Figure 3 suggests that older users are using Facebook to communicate with younger users.

The political use of Facebook is shown in Figures 4, 5, and 6. In comparison to other topics, politics is not a popular item of discussion on the site. Facebook tends to focus on fairly personal topics, with topics on special occasions, life philosophy, and general mood being the most popular topics.

However, when politics does come up, it seems to be focused on presidential elections and candidates, as shown in Figure 5. Again, not surprising. Most people tend to ignore politics except when elections are at their most salient. Figure 6 suggests that as a topic, politics seems to be much more popular with older cohorts. When political discussion occurs on Facebook, it is more common amongst cohorts that are older.

Taken together, these figures may suggest that older users are discussing politics with younger users on Facebook. This informal analysis is not showing causality, but it would be an interesting to use a large-n study to follow up on this hypothesis. Older users are one of the last adopters of Facebook. However, given its extreme popularity with one billion active users around the world, researchers should be more concerned about how all populations interact with the medium.

Pew Research just released the latest round of data on social networking and political engagement. I confess that I have a special interest on this topic since I’ve published a co-authored piece that explores how Facebook affects political participation and political knowledge.

In PEW’s latest report, they identify an interesting contrast between the 2008 and 2012 elections. In 2008, only 26% of people joined a social network of any kind. In 2012, 39% of people used social networking sites specifically for political purposes. That’s a dramatic jump.

In the paper I co-authored, we found that using social networking sites could encourage political participation. Specifically, joining political groups on Facebook would encourage people to do things like vote, volunteer, sign petitions, boycott, etc., In 2012, PEW reports that 12% of adults belonged to a group on a social networking website.

While these 12% of adults may be encouraged to participate as a result of joining online political groups, our study suggests that adults may not be learning anything from this interaction. In survey research, we found that there was no discernible difference between those people who joined political Facebook groups and those who did not.

Our findings are indeed troubling. They suggest that the Internet may not be facilitating political engagement in an entirely ideal way. Nonetheless, future work needs to continue to investigate the ways that the Internet changes our interaction with the political world.

I’m glad to see my extensive knowledge of Arnold Schwarzenegger movies has finally come in handy.