Archive | Social Media

Sharpening Research Questions on Twitter

This is a guest post by Jeremy Pressman.  His previous contributions are here.


While anecdotally many academics are on Twitter, I still hear derisive comments about what one can do with 140 characters. In short: waste of time. So I was struck this morning by the rapidity with which Twitter conversations raised important research questions. Twitter may not be the place to find the fully-fleshed out answer, but as an interface between scholarship and the real world, it seems it can raise important questions. The rapid, interactive nature may help sharpen such queries quickly.

Case #1: Yesterday, Amnon Lipkin-Shahak, a retired military leader, died in Israel. He played a role in many of Israel’s most important historical moments such as the 1967, 1973, and 1982 wars as well as the Oslo peace process of the 1990s.

Today, Brent Sasley lamented Shahak’s death, noting

Sasley’s sentiment is not uncommon as the number of active Israeli military and political leaders with direct experience in Israel’s seminal historical moments dwindles. The current Israeli president, Shimon Peres (age: 89), remains a major exception.

But why does Israel not seem to have a succeeding generation of “great” leaders? More generally, how do great leaders emerge? If by great we mean visionary, willing to take risks, and effective, is there a theory of great leader emergence?

I could see either side as plausible. The definition of a great leader could be one who grooms great leaders to succeed him or her. Or, a great leader could be so powerful that even in a democracy there is little space left for younger leaders to grow. There is a traffic jam of sorts as the same people hang around and make the key decisions as they rotate through key leadership positions. What does the record show?

Case #2: On his blog, Dan Drezner asked, “When is Obama gonna try on foreign policy?” Noting the paucity of treaties passed under President Obama, Drezner highlighted a Monkey Cage post and the idea of opportunity costs. In the Monkey Cage post, Erik Voeten wrote: “The Senate’s agree and consent process takes away legislative time and political capital that could be used for other, perhaps more valuable, legislation.”

On Twitter, Shashank Joshi strongly endorsed Drezner’s post:


After some back and forth with Joshi, I began to wonder. An opportunity costs argument suggests the president chose to spend his time on something other than pushing treaties or going to the wall for Susan Rice. What was he spending his time on? In short, it is easy to say the president should spend more time on something, but can you also tell me what he should have spent less time on? (A day is still 24 hours)

Moreover, how do we know how a president or any leader spends his or her time? Is there a measurement unit of presidential (or executive) time or political capital? I would sure love to see a comparison across administrations.

Case #3: Back to Israel. Israeli voters will select a new parliament on January 22, 2013. Polls suggest that the party of the current leader, Prime Minister Benjamin Netanyahu, will get the most seats and thus he will likely be prime minister again.

If so, current polls also suggest he will face a choice. He will, as Israel prime ministers have always done, need to form a coalition government to get at least 61 seats in the 120-member Knesset. Will he turn to the right and parties like a Jewish Home or to the center-left and parties like Labor, Livni, and Yesh Atid?

In a Twitter discussion, I wondered if he would choose parties based on a policy choice: Would he want to renew negotiations with the Palestinians or not? If yes, choose center-left parties. If not, choose right and religious parties.

But Abe responded with personality, not policy:

Could a bad relationship between Netanyahu and both Naftali Bennett (Jewish Home) and Tzipi Livni shape the next coalition government more than policy and interests? Would Netanyahu choose or avoid party partners based on whether he got along with their leader? The discussion is reminiscent of the obsessing over the Obama-Netanyahu relationship and how it might or might not shape US-Israeli ties.

More broadly, what factors affect party choice during coalition formation in parliamentary systems? Is the key interest, personality, policy, or some other factor?

In all three cases, the interactive nature of Twitter leads to free-flowing discussions that often – not always – allow for the refinement of questions and competing evidence. For those of us who benefit from an intellectual community, twitter can be a useful source for generating questions.

My point in highlighting these three cases is not that these questions are unanswered in the existing literature. But twitter may push people in the direction of being curious and seeking out what has been written. They may well find out what the existing answers are, if any, and then, if a puzzle or gap remains, pursue it in a traditional academic fashion. Or at least blog about it.

Continue Reading

Tweeting Tragedy: A First Look at Newtown Related Hashtags on Twitter

One of the ways people have responded to the horrors of Newtown has been through the use of social media. I don’t know if my experiences are typical, but I actually learned about Newtown from posts on Facebook, and over the ensuing hours my Facebook feed was filled with responses to the tragedy. My initial impression was that what started as an outpouring of grief and anguish began to crystallize into anger at the status quo, which in turn gave way to calls for political action.

Moving from the anecdotal to the more systematic, here is a list of the number of tweets featuring prominent trending hashtags and terms in the hours after the shooting. The data were collected starting at 7:00 PM EST on Friday and run through the end of Monday, December 17th.

Upon first glance, we can see that while there are over 700,000 tweets that include either the hashtag #prayfornewtown or some version of the town’s name (e.g., the singer Nick Jonas tweeted “My heart is heavy today hearing the news about Newton, CT. Praying for all the family and loved ones e ffected by this”), there are just as many tweets specifically referencing either the NRA or gun control (e.g., BBC newscaster Andrew Neil tweeted “If brave teachers can stand up to a gunman – and lose their lives in the process – why can’t US politicians stand up to the NRA?”). If we expand the political terms to include “guns” (e.g., commentator Charles Garcia tweeted “Gov’t regulates ladders which kill 300 Americans a year, more than guns, which kill 30,000!”) then the number of (potentially) political tweets in the sample significantly exceeds those that simply refer to Newtown.

Perhaps more illustrative is to look at the use of these trends over time:

Two trends are worth noting. First, apropos of this previous post on The Monkey Cage about declining attention to gun tragedies, use of all of the terms drops dramatically after Americans go to bed on Friday, and never returns to the levels to which they originally rose. It is interesting to note that after a lull on Saturday, tweets begin trending upward again Saturday evening, throughout the day on Sunday, and then again during the day on Monday. With the caveat here that we are only looking at trends over three and a half days – the data referenced in The Monkey Cage post is over the course of weeks – it does seem that Twitter offers another useful window into public attention to these types events. If so, then we should be able to use Twitter going forward to learn if public attention can be sustained over a longer period of time now as compared to previous tragedies of this nature, or, to put it another way, if this time will be different.

Second, and perhaps more importantly, the data do seem to confirm my initial anecdotal impression of emotional responses morphing into more political ones. By far the most popular hashtag on the day of the shooting is #prayfornewtown. However, as we get late into the day Sunday and throughout Monday, “gun control” and #NRA emerge as more popular Tweets. While we of course need to get a better understanding of the content of these tweets before we make any definitive conclusions, at first glance this seems consistent with a picture of initial emotional responses giving way to a call for more political responses. This observation is consistent with a growing literature on emotion and politics that suggests that anger and fear can motivate political participation. Whether political action is forthcoming in the aftermath of Newtown remains to be seen, but perhaps following Twitter in the coming weeks can provide important clues as to whether this will or will not be the case.

[Figures and tables by Pablo Barberá using data from the Social Media and Political Participation Lab at New York University.]

Continue Reading

The Unstoppable Rise of Social Media

Social networking has spread around the world with remarkable speed. In countries such as Britain, the United States, Russia, the Czech Republic and Spain, about half of all adults now use Facebook and similar websites. These sites are also popular in many lower-income nations, where, once people have access to the internet, they tend to use it for social networking.

The above quote is from a recent Pew Global Attitudes Project report drawing on surveys from 21 countries this spring. Here’s the breakdown of social media usage by topic and country:

Lots to chew on here, but can’t help but immediately notice the prevalence of political usage of social media in the Middle East and North Africa.  Not only are the proportions of users of social media discussing politics higher in these countries than anywhere else, but Egypt, Lebanon, and Tunisia appear to be the only countries in the study where social media is used more often to discuss politics than sports.

The full report is available here.

Continue Reading

Most emails to parties and candidates across Western democracies go unanswered

There has been a great deal of discussion in the aftermath of the US election on the way in which the Obama campaign stayed in touch with its supporters electronically.   Indeed, the quantity of email sent from the Obama campaign reached sufficient levels that the The Daily Show labeled it “Obama Spam”.  Of course, email can go both directions. With that in mind, we present the following guest post from University of Bologna (and NYU-Florence) political scientist Cristian Vaccari.


In spite of all the hype for social media and sophisticated database technologies, email is still one of the most universal tools for online campaigning. For instance, 92% of American adults who are online use email. Among the many ways in which citizens can contact parties, candidates, and elected officials, email is probably the easiest and most accessible one. Moreover, in light of the widespread low levels of trust in parties and institutions across Western democracies, email might help politicians reconnect with citizens. Yet, according to my research, almost two-thirds of emails that reached parties and Presidential candidates during national elections went unanswered.

As part of a broader study of the online presence of parties, party leaders, and Presidential candidates in Australia, France, Germany, Italy, Spain, the UK, and the US, I tested whether and how rapidly their staffs responded to two types of emails (sent from separate fictitious accounts in the official language of each country): one asking for their positions on taxes (a cross-cutting issue that should not strongly differentiate between different types of parties), the other pledging to be willing to volunteer for them and asking for directions on how to do so. Emails were sent in the two weeks prior to national elections between 2007 and 2010 to a total of 142 parties and candidates. The results speak volumes to the lack of responsiveness among political actors: excluding automated responses, only one in five emails received a reply within one business day.

In a multivariate analysis that included variables measuring technological development, the socio-political context, and party organizational variables, I found that parties tend to respond more than candidates, that more resourceful parties are more likely to answer the issue question (but not the volunteer pledge, indicating that poorer parties are more careful not to waste opportunities to add volunteer hours), and that progressive parties tend to respond more than conservatives do. (In case you are interested, both during the 2008 primary and general election Barack Obama’s campaign answered both emails, whereas McCain’s did not. Overall, US parties and Presidential candidates were less likely than average to reply to both emails.)

Answering emails is not an easy organizational task. In a national election, hundreds if not thousands of messages flow in every day. They must be dealt with individually, which siphons scarce volunteer time off other tasks, and the potential costs of any mistakes (such as going off message or treating the sender less than kindly) are huge because emails can be easily forwarded to other people, including prominent bloggers and journalists that can generate bad publicity. The findings of my research indicate that most political actors across established Western democracies weigh these costs much more than any potential benefits that might come from timely responding to emails from citizens they do not know. This is consistent with previous research indicating that campaigns value maintaining control over their message more than interacting with voters. Recent news reports on how the Obama campaign used sophisticated database techniques to carve out specific niches of the electorate and reach them with highly personalized messages confirms that political organizations are more at ease when using the web for messaging that they can fully control.


Continue Reading

Science out of Chaos: Network Analysis of #APSA2012 Hashtag

Javier Sajuria, a political science PhD student at University College London, has mapped the social networks surrounding tweets using the #APSA2012 hashtag. His full report (repeatedly being updated with notifications, not surprisingly, through his twitter feed) is available here, but here’s the basic picture (as of 5:49 EST on August 28th):

(And yes, I’m in there (@j_a_tucker) in the pink group, as well as frequent references to our Monkey Cage APSA Poll.) Now please excuse me while I tweet about this post, which I hope many of you will then retweet!

Continue Reading