Follow Us

Sunlight Foundation has had such an incredible journey from 2006 to now. Join us in celebrating eight years of success by viewing and sharing our interactive report, Milestones and Metamorphosis.

Help us build on our eight year        momentum!

Contribute before year's end:
Donate to Sunlight today!

Donate

Listening in on Twitter

by

This is days past the deadline, completely open and written by a non-profit, and has no revenue model associated with it so it isn't intended to be taken seriously as a response to the recent YCombinator Request for Startups. But RFS 3: Building Things on Twitter got me thinking about what I'd like to build on Twitter. Here's the idea:

Government spends a lot of time and money on building tools that allow it to better communicate with constituents. Whether it be the new, flashy White House website in the Obama administration, or the Republican Whip Office's WhipCast(remarkable choice of name), Government is in to building what it would call "cutting edge" applications online. But they traditionally share a common trait-- they're all new, and interesting ways for government to be able to broadcast information to you.

On the other side of the conversation, citizens have been spending a lot of time figuring out how to do the opposite-- how to send information back to government. The Open Forum Foundation has been spending a lot of time on this. With frequent Sunlight Labs contributor Jim Gilliam, they made GuvLuv that allows people to send twitter messages to their representatives. There's also the concept that so many keep trying about building a Get Satisfaction for Congress. Lots of energy from both companies and volunteers have gone into trying to figure this out-- how to best send information from citizens back to Congress.

So in short-- lots of technical advancement in terms of allowing each side to amplify their voice in the conversation. But "voice" is only one half of a conversation, there's also listening. And the only "listening" technology I know of that Congress uses regularly and pays attention to is called "polling."

I think there's an opportunity, especially with a location-aware web, for Congress to start investing in new technology that helps them listen better. Knowledge as Power proposed basic advocacy headlines as a structured way for people to send emails to legislators in a way that could be easily heard by them, and that's a good step, but I think that technology exists today that would allow for members of Congress to really hear what's going on in their district.

With Twitter adding geotagging on a per-tweet and per-user basis, all the sudden you're able to pull out all the tweets that come from a particular congressional district. On top of that, actually pulling out sentiment on issues isn't that difficult. The vocabulary tends to be constrained, and the number of issues people are talking about at any given time tends to be small. Imagine a view of a congressional district that told members of congress how many times they were mentioned in a positive or negative light by their constituents, or how frequent "Healthcare" comes up in tweets in their area.

Taking it further, you could build a vocabulary of issues like "health care","climate change","bailouts","recovery.gov", extract them from Twitter, along with their geo-locations, and do a sentiment-analysis on them either with software, good old fashioned human beings by using something like TransparencyCorps or MechanicalTurk, or a mixture of both like the folks at PeopleBrowsr.

All the pieces are there, making it really easy for government to go beyond calling people and asking them what they think, and start listening to what they're saying. There are though, a few problems I can think of--

  1. Because the whole system is open and the cost of creating an account is negligible, gaming the system is particularly easy. This is a problem though, that as Twitter grows, will have to get solved.

  2. Another is that geolocation doesn't imply residence. For people, especially in population dense areas, they may traverse several political boundaries on a regular basis.

  3. Twitter doesn't quite contain a representative sample of any population. Though again, as Twitter grows, it may. And twitter profiles do contain, often, some basic demographic information that could be auto-obtained.

Those are, amongst a plethora of others, some of the bigger problems of the system. I'd argue though that Twitter is just one of the solutions too. The over-arching thing here is-- hey Congress, instead of investing in tools that allow you to tell us more things, how about investing in tools that help you listen to us more?