One of the first social studies lessons I remember, from elementary school, was about representative democracy. (This was shortly after the lesson where I learned that Philadelphia is shaped like a woman’s head viewed in profile – think large bun and poofy bangs. But I digress.) In a representative democracy, we, the people, elect Representatives, who then represent our interests. This representation is based on geography. And yet, as we recently learned, it turns out that the situation is really not so simple. Legislators may be getting their votes from inside the districts they represent, but it turns out that the money that supports them often originates elsewhere. Why is it important? Because it means that elected officials who raised a majority of their campaign funds from outside their district, might end up having ties to a community that has not elected them. So what time and attention would these Representatives have left to address interests of the voters they represent?

This disparity was brought home quite concretely by our recent collaboration with MAPLight.org analyzing the geographic sources of campaign contributions to members of the U.S. House of Representatives. MAPLight.org is a nonprofit, nonpartisan organization that has assembled a vast public database to illuminate the connection between campaign donations and legislative votes (the acronym MAP stands for ‘Money and Politics’). Having heard of Azavea’s work on our Cicero™ legislative district boundary and elected official API, MAPLight.org approached us to assist with the mapping component of a study they were undertaking to analyze the geographic sources of campaign contributions to U.S. Representatives.

Map displaying dispersment of campaign contribution sources to a state legislator.

Before we could get to the mapping component, we had to first figure out the actual origins of nearly a million campaign contributions by geocoding them and locating them within congressional districts. In an effort to be as precise as possible, we chose to use only contributions that were matched at the address level. For all other contributions, we checked the Zip code for overlap with congressional districts. The creators of district boundaries, of course, do not take Zip codes into account when creating their districts, so this process left many contributions with two or three (and sometimes more) possible districts. Feeling optimistic, MAPLight.org chose to give the legislators the benefit of the doubt – if the ZIP code of the contribution overlapped at all with the congressional district, it was counted as an in-district contribution.

Percentage of funds raised from out-of-District for U.S. House members.

While MAPLight.org did the actual data analysis, the next step for Azavea was to find a way to visualize the results. The vision was to create a map for each Representative showing the relative amounts of contributions originating in each district across the country. But with 421 legislators in the study, creating each map by hand was not an option we were eager to pursue. Our first step was to take advantage of ESRI’s ModelBuilder, creating a model that would quickly update the map based on an input legislator. But while ModelBuilder is a great tool for automating geoprocessing tasks, it was only able to solve half of the problem – we also needed to export each map to an image file. For this, we turned to ESRI’s ArcView‘s built-in scripting capabilities using Visual Basic for Applications. By combining the two features, we were able to rapidly produce 421 maps for the report, and visually back up MAPLight.org’s conclusion that U.S House members raise 79% of their campaign funds from outside their districts!

The maps can now be found alongside the full report on MAPLight.org’s website.