Visualizing Network Data to Keep Corporations Accountable

Facilitated by Skye Bender-deMoll

Session Description

Skye will demo www.dirtyenergymoney, which is a new tool for visualizing campaign contribution relationships between oil companies and the US congress, and then lead into a collaborative discussion of network data tools and other possible applications of network mapping in other fields.

Presenter: Skye

Activity: place your hand on the shoulder of the person in the group you’ve known the longest. [Result is a network map; result is relational information] Contrast with other ways of organizing information about people, like the "spectrogram" opinion exercise at the beginning of the conference.

You can get relational data from social network sites (eg facebook). Facebook uses a social network to produce a social networking site.

Touchgraph.net has a facebook app that can produce a network map of your fb friends. http://apps.facebook.com/connectedbook/ There is emergent information that comes from the network that isn’t necessarily tagged anywhere. It can reveal groups, and linkages between groups

Network lingo: People = “nodes,” links = “ties” or “edges”

But "nodes" don't have to be people, they can be companies, Political groups, concepts, etc.

Example of map of membership relations between groups discovered by survey of delegates at the major party conventions: http://skyeome.net/wordpress/?p=377

http://Dirtyenergymoney.com maps the connections between oil companies and politicians. You can explore the network of donations (total number in the millions) and even get the relevant FCC documents. This is a great tool but as a political outreach tool Skye questions its utility. The people using it are the choir. It also requires some orientation to learn to use it (like any tool). Here is a blog post with more background (including an animation of the network): http://skyeome.net/wordpress/?p=220

http://Prop23.dirtyenergymoney.com explores the funding connections between California’s Prop 23 and Prop 26. Interpreting the graphic: $ flows left to right. Notice how the $ often flows indirectly. CA has stronger disclosure laws so you can track these donations. On the federal level there’s much less transparency.

You need the stories to help the public make sense of these strange graphics. Narrative is powerful. These complex infographics are still much more intelligible than the raw data record of the donations.

Many of the barriers are on the user interface side. The technical capacity exists to display even more info but it’s overwhelming.

These tools may be more useful for strategy and campaign planning than for the general public. Robyn suggested some sort of Prezi zooming presentation http://prezi.com/ tracing the data on the site. You’re telling the story but not limiting the display (or potential) display of data. Limits undercut your credibility.

Bruce works on campaigns and thought the Prop 23 graph would have been really useful back in June. This work is only as good as the campaign finance disclosure laws. Most of the info is available after the election. Many places don’t even digitize their records. The better we get as using this data the more process loopholes get inserted by opponents.

Some other examples:

Blog and political reporting sphere is a sociologists wet dream. A French org made a graph of the US “politicosphere.” http://PoliticoSphere.net Good UI aspects: you can zoom in, more connections revealed. Cons: once the graph get complicated it’s hard to interpret.

Orgs on the left and the right are trying to standardize nonprofit 990 data to illustrate connections. Skye shared an example of a network graph that he created with an open source package called GraphVis. Interesting data, but map is still too complex to be useful: http://skyeome.net/wordpress/?p=412

Data that comes out fast is news. Data that comes out more slowly can be used for coalition and campaign analysis/strategy. Uncover the infrastructure. Improve organizer “powermaps” (process of analyzing allies/opponents/persuadables). There are some differences in terms of inputs (network graphs input quantifiable data, power map process has organizers provide their feelings about where groups fall on the chart).

The network map by itself often isn’t that interesting, it becomes interesting when you combine it with other categorizations and background info.

Other useful links:

Skye's report to AAAS on possible human rights uses of network maping. Includes a plain language intro to networks, as well as many of the examples demoed: http://shr.aaas.org/networkmapping/

VisualComplexity.com, a great visual resource of network-ish maps and projects: http://www.visualcomplexity.com/vc/

A network movie illustrating a model of HIV spreading across sex contact networks: http://csde.washington.edu/statnet/movies/index.html