Murphy's Law: Building an Open Source America with Open Source Data

Posted 07/24/2009 at 8:00am | by David Murphy

3

Comments

Print

A coalition of some of the biggest names in the OSS world have banded together to create Open Source for America, a brand-new advocacy group that's going to try and highlight the advantages of open-source software to help achieve the goals set out in President Barack Obama's push for an open-data government. But as we pause to "ooh" and "ahh" at the list of companies and open-source celebrities contributing to the new group--Novell, the Mozilla Foundation, the EFF, Tim O'Reilly, and Mark Shuttleworth, amongst many others--let us not forget the uphill battle that the concept of "openness" tends to face in the government sector.

I'm not suggesting that the OSA's cause is anything but noble: Instead of pitching open-source as a replacement to proprietary code, the group wants to level the playing field and allow both to equally compete for governmental contracts. That does have practical implications for the common person. Increased savings and reliability will allow resources spent for upkeep to be shifted to additional needy sources, and open-source software should--in theory--be easier and quicker to deploy than proprietary code given the larger user base that could feasibly assess and contribute to such a project. When it comes to efforts like standardizing a way to share a patient's health records across a wide swath of federal, state, local, and private organizations, multi-tiered cooperation and speedy results are a good thing.

Nevertheless, I just can't find myself getting that excited over open-source software when we still have fundamental issues of transparency and openness in governmental data. There's a wealth of information out there that's free and easily accessible to the public. But that doesn't mean that legislators, agencies, and departments are going out of their way to make this information as useful as it could be. In fact, it was only as recently as two months ago that the U.S. Senate opened up its own voting records for third-party applications and mashups.

Open Data's in the House

The House of Representatives and Senate have both used XML to exchange legislative documents since 2000. It's pretty easy to find XML feeds for the bills passing between the two legislatures via The Library of Congress's THOMAS database--the central information repository for all legislative activity in Congress. Pulling up an XML listing of how the votes fell out in the House is also simple, although you'll have to jump to the Office of the Clerk for that information. The Senate, not-so-much. Up until May, when a formal request by South Carolina Senator Jim DeMint was finally accepted by the Senate Rules Committee, this information was restricted to straight-up HTML feeds.

Why is any of this important? Because third parties wishing the use the data to track things like voting records and attendance, filibustering, and comparisons between legislators would have to input this information manually in order to get an accurate database for external applications or data mashups. These amalgams, in turn, offer increased access to the legislative record for your average, everyday American who has no time to scan over arcane THOMAS listings to discern the status of a piece of legislation or, come election time, a Congressperson's legislative activity.

Representative Melissa Bean, Democrat of Illinois, put it best in an interview with Politico. “Coming out of the business world, I think results matter. How can people really track results if they don’t even know how their representatives are voting?” she said. Bean, in turn, has introduced legislation that would compel the Office of the Clerk to create an electronic list of all votes taken in the house organized by member. This feature, currently lacking in the XML feeds for roll call votes, would allow visitors to quickly and easily discern exactly how and when a particular member has voted throughout the congressperson's tenure.

Free Data Federalism

And this is just the situation in the U.S. Congress. Check out Data.gov, the brainchild of the Obama administration's first-ever CTO, Vivek Kundra. Specifically, head on over to the "state/local" portion of the site to see just how many states in the union have begun to offer up official archives with machine-readable datasets (like XML) for public consumption. I'll spoil the answer for you. Not counting the District of Columbia, there are two: California and Utah. To be fair, there's a varying degree of open information accessibility in the individual state legislatures--New York, for example, has pushed past RSS-feed-like XML feeds and offers full APIs for data access. It's a move that many wish to see enacted at the federal level.

In a perfect world, the results of our representative democracy would be available for perusal by any member of the public. But when I say available, I don't just mean locked away in some complicated, searchable database. Let the third-party world of application developers, information enthusiasts, and political junkies sift through the deliverables and create their own compelling derivatives for public consumption. Based on the impressive results that have come from the laborious combination of automatic and manual data entry, I can only imagine the kinds of benefits that an average citizen could acquire as a result of a truly "open" government. Open-source software is a great discussion point. Let's throw a bigger bone to open-source data, too.

A few data-crunching sites worth surveying:

Govtrack.us

Fillibusted

e-PaperTrail

OpenCongress

Maplight