Deen Freelon, American University
Abstract
The Arab Spring has furnished a prime set of test cases for research on the major uses of digital media in general, and social media in particular, in protest and collective action contexts. Along with a number of articles employing traditional social science approaches, several big-data studies of social media have already emerged (Howard et al., 2011; Lotan et al., 2011; Papacharissi & Oliveira, 2012; Starbird & Palen, 2012). These studies have invariably focused on the producers of social media content, completely ignoring the much larger population of consumers of said content. Studying patterns of online message consumption around momentous collective action events is critical in ascertaining the overall utility of digital tools to the actors who participate in them. Two aspects of this overarching research program stand out as particularly pressing. First, the question of where the consumers are located bears directly on broader theoretical questions of social media’s role in directly abetting collective action and citizen journalism. If most consumers are outside of the affected countries, the case for social media’s role in directly facilitating protest is accordingly diminished. Second, examining whether the most-consumed content originates from traditional media sources (e.g. the BBC, CNN, the New York Times) or from social media platforms accessible by anyone (e.g. Facebook, Youtube, Tumblr) speaks to theoretical concerns about the degree of control large news organizations still maintain over the framing and distribution of protest-relevant information.
To investigate these two research questions, we rely on data drawn from Twitter, a popular microblogging service; and bit.ly, a popular link-shortening service frequently used with Twitter. We began with four non-exhaustive full-text hashtag archives, one for the following four MENA (Middle East/North Africa) countries: Bahrain (#feb14), Egypt (#jan25), Libya (#feb16), and Tunisia (#sidibouzid). These archives were collected through the service TwapperKeeper, which prior to March 20, 2011 allowed users to export any of its vast collection of running tweet archives free of charge. All told, the four hashtags used in this study contained 1,677,997 tweets spanning roughly from mid-January 2011 through the end of March 2011. We extracted over 95% of all bit.ly-shortened links found among these tweets, including those administered by bit.ly but bearing a branded domain (e.g. “nyti.ms” for the New York Times). Our analysis incorporates all clicks received by the 44,805 unique links we extracted from the hashtag tweets. We obtained an array of metadata directly from bit.ly pertaining to each link, including the number of times it was clicked, the precise second each click occurred, and the city-level location of each individual who clicked on it.
Overall, we found that click patterns were highly event-driven, spiking on days of major significance such as January 25 (Egypt’s first major protest day) and February 11 (the day Hosni Mubarak abdicated). Between those relatively rare days, click rates fell into a muted pattern of very small peaks and valleys. To address the question of where the consumers of the linked content were located, we classified each city identified within each hashtag into one of three categories: within the country to which the hashtag was devoted, within the broader MENA region (operationally defined as Algeria, Bahrain, Djibouti, Egypt, Iran, Iraq, Israel, Jordan, Kuwait, Lebanon, Libya, Mauritania, Morocco, Oman, Qatar, Saudi Arabia, Somalia, Sudan, Syria, Tunisia, United Arab Emirates, and Yemen), or elsewhere in the world. Within all four hashtags we found that tweets from outside the MENA region vastly outnumbered those within it. In Egypt, Libya, and Tunisia, over 75% of clicks came from non-MENA countries; that percentage fell to 62% for Bahrain, which is by far the most wired country we examined with a 57.1% net penetration (Miniwatts Marketing Group, 2011). On the basis of these results, we conclude that Twitter was not a major source of Arab-Spring related information for large numbers of people in the countries we studied.
Answering our second research question about the relative prominence of traditional and social media content within Twitter required us to classify each link domain as belonging to one of the two categories. We were able to positively classify over 77% of all our 44,805 bit.ly links under one of 40 traditional media or 19 social media domains. (A complete list of these domains is available in the full paper.) Within this subset, 71% of all clicks led to a traditional media domain, while 29% led to social media domains. Traditional media also constituted a majority of the top five most-clicked links in each hashtag. Taken together, these results constitute strong evidence that traditional media still play a significant role in distributing news and information concerning international protest activity. To be sure, much of this content likely represented ordinary citizens sending text, photo, and video dispatches from their mobile phones. However, this content was largely not shared directly between citizens—rather, traditional media organizations served as a distribution hub for it, as has long been the case.
In summary, we find that links posted to MENA hashtags in the early days of the Arab Spring were visited mostly by non-MENA audiences who overwhelmingly viewed traditional news content. This interpretation supports the notion that when countries with low Internet penetration experience protest activity, social media’s primary role is as a megaphone to relay relevant news and opinions to the broader world.
References
Howard, P. N., Duffy, A., Freelon, D., Hussain, M., Mari, W., & Mazaid, M. (2011). Opening Closed Regimes: What Was the Role of Social Media During the Arab Spring? Project on Information Technology & Political Islam.
Lotan, G., Graeff, E., Ananny, M., Gaffney, D., Pearce, I., & boyd, d. (2011). The Revolutions Were Tweeted: Information Flows During the 2011 Tunisian and Egyptian Revolutions. International Journal of Communication, 5.
Miniwatts Marketing Group. (2011). Africa Internet usage, Facebook and population statistics. Retrieved from http://internetworldstats.com/stats1.htm
Papacharissi, Z., & Oliveira, M. (forthcoming). Affective news and networked publics: The rhythms of news storytelling on #egypt. Journal of Communication. doi:10.1111/j.1460-2466.2012.01630.x
Starbird, K., & Palen, L. (2012). (How) will the revolution be retweeted? Information diffusion and the 2011 Egyptian uprising. Proceedings of the ACM 2012 conference on Computer Supported Cooperative Work (pp. 7–16).