From Madison to Occupy Wall Street, the economy led as the most covered news story in U.S. news coverage for 2011, according to Pew Research Center’s Project for Excellence in Journalism. Much of that attention was driven by protests against growing income inequality here in the United States, as well as globally.

For an international communication seminar project last semester, I wanted to examine how social movements use information communication technologies to challenge the dominant economic order. To do so, I chose to compare Occupy Wall Street with the Spanish 15M, or Indignados movement, which may be less familiar to a U.S. audience.

On May 15, 2011, one week before municipal elections, thousands took to the streets in 50 Spanish cities to protest corruption and demand “real democracy,” calling for crisis management by “the people and not the banks” (see Periodismo Humano). Forty demonstrators gathered in Madrid’s main square, Puerta del Sol, into evening of May 15 talking about the country’s future decided to stay. In the early morning hours of May 16, those in the plaza made a key tactical decision to negotiate with police over their presence, who allowed them to stay the night. And they camped, sparking what is now know as the 15M movement, for the date of its commencement, or #SpanishRevolution after one of the movement’s main Twitter hashtags.

As events of the past year illustrate, while the social problems associated with economic globalization are correspondingly globalized, protest actions such as Occupy and 15M are place-based.

Network Structure
In a mixed methods approach, I followed the approach of Bennett, Foot and Xenos (2011) using qualitative think description to inform social network analysis of the structure of the Spanish 15M and U.S.-based Occupy movements, as well as their linkages.

I hypothesized the relationship between narrative and structure along the lines of: narrative allows us to see the values that underlie how activists use technologies, in an iterative process this then informs what we learn about how they are using technologies to mobilize supporters. Taking this approach, we are better able to understand why activists make the technological choices they do. Looking then at the network structure allows us to visualize the national and global networks.

The two movements share many values, such as an emphasis on the process of constructing a new society through participatory decision-making and working group structures. Yet under pressures from law enforcement and arguably in reaction to a mainstream media focus on its early lack of demands Occupy Wall Street developed a trajectory of actively denouncing demands. This opens possibilities for more diffused actions and less centralization within the network.

I tested a series of structural predictions (please contact me if you would like a copy of my working paper) to visualize the ways in which the 15M and Occupy movements link to each other via activist websites. The data depicted here was collected on December 11, 2011 using the program Issue Crawler from the Govcom.org Foundation at the University of Amsterdam. I set the program a one iteration web crawl, going two pages deep in each of the sites. In order to be included each network, an actor needed to be “colinked,” i.e. have links to or from at minimum two of the URL starting points for that network.

15M Network
The Spanish 15M network is based on 72 starting points from the coordinating Toma la Plaza’s listing of local and global movements. As a colink visualization, it includes websites that are linked (inlinks or outlinks) to at least two sites in the original listing of 15M groups. I used the open source data visualization program Gephi to visualize the data.

I experimented with the data visualization functions in Gephi. For this sociogram, the nodes are scaled to the number of inlinks (in-degree) and the edges (lines indicating links) show the target of the link (or what is linked to by a source).

Occupy Network
The U.S.-based Occupy network is based on 87 starting points from the coordinating Occupy Together listing of local occupy groups on its “Actions and Directory” page (excluding Facebook pages). Note that this network includes groups based outside of the United States, such as Occupy Italy and Occupy Amsterdam.

For this sociogram, the node size is scaled to the number of inlinks and the edges are the target (but not showing the strengthen of the tie as is the case for the Toma la Plaza data).

Global Change Network
A depiction of the global linkages between the two social movement spheres in Spain and the United States, as rendered by Issue Crawler is based on 21 starting points from listing of main activist organizations (available here), along with the “actions” or “events” pages on each site.

For this sociogram, the nodes are not scaled to degree and the edges show link targets proportional to number of inlinks.

Findings

National Networks. Neither network proved to be particularly dense. The 15M Toma la Plaza network has a directed density of 0.070, while the U.S. Occupy Together network has a directed density of 0.032. The 15M network also displays a higher level of clustering, with an average clustering coefficient of 0.126. The average clustering coefficient for the Occupy network is 0.055.

In terms of eigenvector centrality, for the 15M network the social networking site Twitter has the highest eigenvector value of 1, while the coordinating site Toma la Plaza has a much lower value of 0.530. “Eigenvector centrality” is a measure, similar to Google’s PageRank, of how connected a node is to other well-connected nodes within a network (Hansen, Shneiderman, & Smith, 2010, p. 41).

Two other key actors have low values: Democracia Real Ya! (0.348) and the Madrid Toma la Plaza (0.376). In contrast, for the Occupy network Occupy Together has an eigenvector centrality of 1. The only other two sites with values of more than 0.75 are Twitter (0.940) and OccupyWallSt.org (0.818). In contrast, 91 sites in the network have an eigenvector value of 0. On the whole, eigenvector values in the 15M network are lower and more evenly distributed, suggesting a more horizontal network. However, this finding must be qualified by the fact that 15M has been in existence for twice as long as the Occupy movement, giving activists more time to develop coordinating structures of the kind that are only now emerging in the Occupy case. It is clear that Twitter serves as a crucial platform for information exchange in both networks. Other new media technologies present are Livestream and Facebook, as well as the web platform WordPress.

Global Network. Not surprisingly the site within this network with the highest number of inlinks from the network is Twitter, again indicating its importance as a forum for information exchange. Interestingly, the second highest site in terms of inlinks from the network is a website translation service ICanLocalize. Its eigenvector value is also 1. Also ranked highly is the website of the WordPress Multilingual Plugin. There also does not appear to be a substantial amount of inter-linkage between linguistic spheres beyond the coordinating sites. The network’s directed density is 0.042 and the average clustering coefficient is 0.075. Sites with closeness centralities of 1, include: the “How to Camp” of Take the Square, the 10December Day of Action for Human Rights (a project of Take the Square), InterOccupy, London’s New Economics Foundation, the Spanish Democracia Real Ya!, Occupy Oakland, and YouTube.

This research is ongoing and I would welcome any feedback!

Related Research
Here are links to other 15M and Occupy research I find interesting:

Analyzing Newspapers Front Pages

#OccupyWallStreet: Origin and Spread Visualized

The Spanish Revolution & the Internet: From Free Culture to Meta-Politics

Understanding the Occupy Movement: Perspectives from the Social Sciences

Visualising 14 Hashtag Networks #OccupyData

References
Bennett, W. L., Foot, K., & Xenos, M. (2011). Narratives and network organization: A comparison of fair trade systems in two nations. Journal of Communication 61, 219-245.

Hansen, D., Shneiderman, B., & Smith, M. (2010). Analyzing social media networks with NodeXL: Insights from a connected world. Amsterdam: Morgan Kaufmann.