Today I attended the Berkman Center luncheon presentation by Victoria Stodden of Stanford University on “The Internet and Democracy: Problems and Ideas.” (The event was webcast live.) Her talk inspired an animated discussion of what empirical methods are most appropriate for studying the Internet's effect on the democracy. Victoria said that the Internet could potentially affect democracy in three way:

1. by disseminating information (increase in accessibility of information, increased accountability, more informed citizens)
2. as a tool for democratic processes (online voting, activist mobilization)
3. as a means of educating people about democratic values (respect for human rights, freedom of expression)

She then developed this taxonomy with examples, and addressed negative opinions of the Internet's effect on democracy, such as the increase of polarization and the Babel Effect (too much information which is unnavigable to the individual). Victoria then opened the floor to comments.

The first comment came from Berkman fellow Ethan Zuckerman, who questioned the wisdom of studying the Internet's effect on democracy by starting out with a taxonomy and then trying to get the empirical data to stick to it. However, he mentioned in the same breath that there are so few empirical cases of the Internet having a dramatic effect on democracy that it is problematic to start from the empirical level as well, because there are too few cases to be able to extrapolate reliably. He also mentioned that multiple causation was a significant research hurdle. he also noted that political events are influenced by a myriad of factors, it would be quite difficult to isolate the effects of the Internet from all the other influencing factors.

Berkman faculty co-director Terry Fischer took the opportunity to ask Victoria if there were any innovations in statistical analysis that could help us isolate the effects of the Internet from different factors which influence democracy. Victoria responded that, despite new methods, there is no statistical magic bullet to the study of Internet and democracy.

Rob Faris, research director of the OpenNet Initiative, pointed out that much data might need to be collected on a individual level, since the effect of the Internet on democracy comes down to influencing citizen opinions and voting decisions. This struck an immediate cord with me, and I mentioned that when I was researching our case study on the 2002 Korean election the data set that I really wish I had was one that connected Internet behavior to voting behavior.

Terry then asked who carries out exit polling in different countries and suggested that maybe these groups would be willing to ask voters about Internet and cell phone. My fellow research assistant Josh Goldstein suggested that the survey methodology used by Carol Darr and the Institute for Politics, Democracy, and the Internet, to study online political influentials might also be useful to us. I think we definitely made some progress in determining the empirical form of the Internet and Democracy project.

cross-posted on the I&D Blog