Civic Hacking
My hobby-turned-career has to do with the intersection of technology and civics: how to use technology to improve civic education and participation and to improve government transparency and accountability. I call creative uses of technology to address these problems "civic hacking".
Apps
I run GovTrack.us, a website that tracks what's happening in the U.S. Congress [video summary]. The site basically runs itself. See this Wikipedia page on GovTrack that someone (not me!) thought to make, and GovTrack's press coverage. GovTrack launched in 2004 and was the first website to apply the principles of open data and Web 2.0 to the U.S. Congress. It catalyzed the development of a community of like-minded developers and influenced the open government movement that we see today. In 2010 it expanded to GovTrack Insider with a small team of paid reporters.
I also had a secondary role in the development of FlyOnTime.us which analyzes FAA flight data to determine how late airline flights tend to be (not as bad as you'd think).
Also see my HackingCongress.org and oGosh Facebook group.
Data
The application nature of GovTrack is just half of the story. What was particularly innovative and really changed the future of open government in this country was that I made all of the legislative data that I was collecting available for anyone else to reuse for free. This data and other data that I've collected is a significant part of what powers a bunch of other websites now. In particular, if you think Sunlight Foundation is the leader here, keep in mind that their grantees MAPLight and OpenCongress, almost half of the entries to their first Apps for America contest including its winner Filibusted.us, and a finalist in their Apps for America 2, ThisWeKnow.org (subsequently referred to on the Senate floor), all rely significantly on data that I put together. (I also co-wrote Sunlight's first big report, The Open House Project Report, while we're on the subject.)
Government Data Policy
But I also work on the policy angle: what should governments do to give the public the best access to its data? I co-wrote/co-led the The Open House Project and The Open Senate Project, run out of Sunlight Foundation, which made recommendations to the House and Senate on how to use technology better in the interests of transparency. We had a press conference in the Capitol building back in 2007 [video]. (Slashdot) In 2007 I was at the workshop that lead to the 8 Principles of Open Government Data which has been influential in its own way (e.g. NYC open data bill, NY state open data bill). I later wrote my own monograph on the subject extending it to some 16 principles.
I've been lucky enough to affect policy at the federal level:
My main "ask" in the Open House Project report was to have the Library of Congress share with the public its database of legislative information, so websites like my GovTrack and researchers can use the data for new purposes without having to gather it in roundabout ways. In November 2007 the House of Representative's Committee on House Administration asked the Library of Congress to look into making their raw database of legislative status directly available to the public, thanks to Congressman Mike Honda (and his staff)'s support for this part of the report. In June 2008 Honda submitted two paragraphs for the House legislative branch appropriations bill asking the Library as well as the Government Printing Office to look into this, which was passed in March 2009 (as part of the House report accompanying the FY09 omnibus appropriations bill HR 1105 (111th Congress)). (Slashdot) We're waiting on action from the Library.
Many of the recommendations of the Open House Project report were turned into legal language by Sunlight's staff in 2008 (here, but defunct) and that was later introduced by Rep. Quigley as H.R. 4983: Transparency in Government Act of 2010, or informally the omnibus transparency bill.
GPO claimed they took the language seriously in the FY09 appropriations report and formed a task force. I worked a bit with Carl Malamud on buying the Code of Federal Regulations from GPO (buying the law!) but as Carl wrote back to the GPO, the files we got were not even worth the money we put in. (As he says, if it's going to be sold, they ought to at least provide a certain level of quality.) In any case, GPO responded to the pressure from our work through Honda's office, from Carl, and from the White House's open government directive and began publishing the Federal Register in XML for free in October 2009 (better backstory), which in turn lead to the redesign of the government's official Federal Register website in July 2010. GPO also began publishing the Code of Federal Regulations for free in XML in December 2009 (promise | statement). There is more the GPO hasn't freed yet. But H.R. 4858: The Public Online Information Act of 2010 (POIA), if passed, gives GPO yet another push. POIA was introduced with the help of Sunlight Foundation, and it calls for all three branches of the federal government to work with an advisory committee on data publishing. (WaPo endorsement)
Something else I had been pushing had been getting the Senate to publish its roll call vote records in XML, as the House had done for many years. As with the legislative database above, this helps the public do new and interesting things with the information, like keeping the public better informed. In December 2007, Senator Lieberman made a public (though cryptic) statement about how to publish Senate votes online, which was because my push for this trickled up the policymaking ladder. In April 2009, Senator DeMint circulated a "Dear Colleage" letter which I helped draft asking the Senate Rules committee to revise their policy on publishing roll votes as XML. Within a few weeks the Rules committee reversed their position and XML came online for Senate votes. (Politico 1, 2, 3; Columbia Journalism Review)
The best practice recommendations for executive-branch agencies regarding data incorporated some of my suggestions in June 2008 (suggestions drawn primarily from the thoughts of many others compiled into the Open Government Data Principles).
The Semantic Web and Linked Open Data began to get more attention in the government community in 2009. I'd been tinkering with that idea since 2005. The man who invented the web as we know it, and who has been pushing for the use of the Semantic Web, encourages governments to take it seriously now too (and cites GovTrack). We'll see where this goes. I'm not really sure it's time yet for the Semantic Web inside government.