Thursday, November 13, 2008

Twittering the election… and wondering if this is the right tool

Many of my friends who are following the US election intensely are supporting Twitter Vote Report. It's a very cool mashup designed to let people report voting irregularities by sending a message to #votereport on Twitter and using a restricted syntax to report on the experience. The website will visualize the reports as they come in and will be able to store reports of slow voting sites and polling places that experience complaints of malfunctioning machines or people preventing voters from accessing the polls.

A sample report:

Syd Sallabanks: #votereport #early Boise 83716 zip. #good experience to vote early. Boise friends follow

I can't help comparing this laudable project to some of the projects I've seen in African countries designed to increase voter transparency. Some of those projects have used SMS. Election monitors in Nigeria used Kiwanja's SMS gateway, FrontlineSMS to monitor the recent presidential election. And SMS likely helped the opposition MDC insist that it had won the first round of presidential balloting this year - election reports were posted outside each polling place, and MDC activists used SMS to report each tally to a central office, where they were tallied and revealed an MDC victory, if not a majority.

But the most effective vote monitoring projects I've seen are in countries with a free and thriving indepedent media. In Ghana, talk radio is by far the most important medium for discussing politics. During the 2000 elections, citizens who had trouble at the polls - groups trying to intimidate voters or prevent some people from voting - called talk radio stations and reported their troubles. This was probably more effective than calling election officials or other authorities - since the obstacles to voting were reported live, the radio stations could continue reporting on the situation until authority figures intervened and ensured people could vote. (It's possible that election authorities might have ignored calls to their offices and claimed they'd never been received.) As it turned out, the 2000 presidential election in Ghana was peaceful and put the opposition party in power for the first time in decades.

The mobile plus radio system works very well for monitoring for two reasons - it's easy for citizens to use (they just call a radio station, something many of them do frequently to participate in call-in shows) and the reports are immediately available to a large audience (everyone who listens to talk radio, which is, basically, everyone.) I'm not sure that TwitterVote covers the same bases, at least by itself. It's easy for Twitter users, and certainly possible for those who don't use Twitter regularly to participate by texting to a shortcode. But the messages directly reach a fairly small audience - there aren't very accurate numbers for active Twitter users, but Techcrunch estimates the number at under a million, which certainly includes some non-US users. So TwitterVote needs to be thought of as collection mechanism for reports, which can be disseminated through other media.

This, for me, raises the question of why Twitter is the right tool to use for this project. Is it because it's easy to crate mashups around? Because it's the tool-du-jour for the digitally experimental set? Or is it a reflection of how impenetrable mainstream radio and television appears to be for most American citizens? The media that continues to be disproportionately important for most American households is still local, broadcast television news, despite declines in recent years and increase in web usage. When we design tools for election monitoring, are we ignoring local news because we expect it to be uncooperative and impenetrable? Or are we just playing with the tools we know and like, whether or not they're the best way to reach a broad audience?

1 comment:

Qué? said...

I think the problem with the mobile + radio approach is that it is difficult to co-ordinate and get a live picture of what is going on at a collective level.

The radio stations out here all compete with each other and none broadcast one signal nationally. This makes it difficult to monitor and report as quickly as might be the case if all information is being texted to one point, no?