Hack 16. Who Are the Neighbors Voting For?

Fundrace can show you which political candidates and parties have the most support in your areaand which of your neighbors are supporting them.

The financial machinations of presidential election campaigns are not as distant, highbrow, or incomprehensible as one might think, and the web site http://www.fundrace.org/ will show you why. Enter your address and ZIP Code into its Neighbor Search, as shown in Figure 2-8, and you are immediately provided with a vista of the political landscape of your neighborhood. By aggregating and geocoding campaign contribution records, Fundrace allows you to uncover a little more about the pocket depth and political sentiment of your friends, coworkers, relatives, and neighbors.

Figure 2-8. A Fundrace Neighbor Search

Fundrace is the result of questions asked and hacks committed by researchers working at Eyebeam (http://eyebeam.org/), a nonprofit, nonpartisan arts and technology organization in New York City. Fundrace is built on a database of presidential campaign finance records published by the United States Federal Election Commission (FEC). These records include the amount and date of each contribution totaling over $200, along with the name, mailing address, occupation, and employer of the corresponding contributor. With individual contributions to campaigns capped at $2,000, and campaigns raising tens of millions of dollars at a stroke, this amounts to a fair heap of information for anyone to try to interpret meaningfully.

In the autumn of 2003, the financing of the presidential primary campaigns, especially those using grassroots fundraising over the Web, was a prevalent topic in the national media. The content and presentation style of other existing campaign finance web sites, such as http://www.opensecrets.org/ and http://www.fecinfo.com/, did not seem likely to attract the attention of anyone who wasn't actively seeking out this information already. The first iteration of Fundrace consisted of a number of simple statistical rankings and a handful of national fundraising maps, intended to help people draw some distinctions among the wide field of candidates running at the time.

Despite offering a modicum of satire, Fundrace 1.0 didn't reach much further than other sites of a similar nature. By far, the most popular feature of the initial experiment was the national red-versus-blue Money Map (Figure 2-9), in which each county was shaded either red or blue, depending on whether Republicans or Democrats had raised more money there, respectively.

Figure 2-9. A Money Map of the United States, Democrats versus Republicans, by county

Figure 2-10. A Fundrace map of Manhattan

The next step in the right direction came when a Bostonian monthly magazine of some repute asked to publish one of Fundrace's maps. They liked the national red-versus-blue but were also interested in something a little more localized. Finally, it became time to geocode the data. After cleaning up the messier street addresses [Hack #81], Fundrace used MapPoint to plot contributions onto what is probably the most recognized several-miles-square chunk of land on Earth: Manhattan. Aggregating by individual building, as shown in Figure 2-10, while maintaining the red-versus-blue coloring convention, produced an immediately intuitive and stark picture of sociopolitical geography .

What quickly became clear upon making this map of New York was that this data could serve as a lens not only for viewing the different candidates but for looking at the world in general. It also became clear that the principle of locality was as applicable as ever. People are most interested in the data that is physically nearest to them. Thus was born the Fundrace Neighbor Search. With contributors' addresses already geocoded, it was a simple matter of geocoding the query address and sorting by Euclidean distance to find the user's closest contributing neighbors.

As they say, the rest is history. The idea, and its results, were so compelling that within only a couple of days of launching the Neighbor Search and city-level Money Maps, Fundrace.org was an extremely hot web site. Web logs couldn't link fast enough, the site was receiving an onslaught of traffic, and the national media was knocking down our door. From all channels, the feedback Fundrace received covered a broad range of opinions. We received many emails thanking us for the valuable public service performed by the site. We also received numerous requests, none honored, to remove records from our database.

Some people were shocked to discover the political leanings of their friends or relatives. Others were embarrassed to know that their whole office could see whose campaign they had contributed to. The occasional amateur watchdog contacted Fundrace about perceived violations of campaign finance law, and a number of academics, hackers, and GIS professionals requested a copy of our database for their own exploration.

Probably the most contentious aspect of Fundrace was the fact that it revealed the home addresses of contributors. Many considered this a serious threat to personal privacy, but there already existed a number of web sites (including the FEC's) where the same information, as part of the public record, could be obtained. Fundrace is of the opinion that the more people who become aware of the public nature of this information, the better. At the very least, if a large enough constituency of voters were to decide that the risks outweigh the benefits, the FEC's policy could be changed.

In the United States of America there exists a tradition of public access to certain kinds of government records. This access is rooted in the principle of government accountability, a necessary feature of any legitimate democracy. The tradition has, of late, been reinforced by so-called Freedom of Information Legislation (FOIL), and by the ease of disseminating such information via the Internet. Despite this trend, the average citizen still seems largely unable make good use of FOIL and the Internet to understand and participate in his own government. Officially published data and documents are often disorganized, poorly documented, or simply too voluminous to handle with limited skills and resources. Fortunately, this leaves plenty of room for hackers like us, and organizations such as Eyebeam, to do lots of interesting and relevant work.

Fundrace is a proof positive example that abstruse government data can excite and engage people, and that all it takes is a few good hacks.

Michael Frumin