Meaghen Brown, Middlebury College
Finalist, $200

Return to Awards list

Healthcare Map

Initially I wanted to map a political talking point as it flowed through news networks and disseminated down to the level of political consciousness, yet I realized early on that this was going to be difficult to visualize unless I moved into the field of graphic abstraction. This map is my compromise. The data comes from the latest Gallup poll on the healthcare bill (3/15/2010)¹. In this poll, participants were asked to provide verbatim responses about their for or against vote in addition to the typical polling information. This data was logged in an excel spreadsheet available to the public. Using a program called Wordle², which generates graphics of type weights based on the relative amount of times they appear in a body of text, I took the responses for people who voted for and people who voted against the health care bill on a state by state basis and produced 100 weighted graphics based on their responses. I did, however choose to omit the word healthcare from the text simply because it would otherwise skew the weightings of the rest of the words making it difficult to pick out patterns. From there I found USA Contiguous Albers Equal Area Conic projection which I used as a base reference when placing the weighted text. This shapefile was later removed. Finally I manipulated the text in Illustrator fit the general shape of the state from which it was derived. Both the scaling and the weight of each word were preserved.

Here I should insert a few disclaimers about the data used. First, the Gallup poll had no record for Alaska, and I was therefore forced to omit it, which turns the United States into an island on the page. Secondly the data is sourced from respondents to a poll, meaning it comes from a limited selection of the whole. Similarly the number of respondents is not equal for each state. I represented this by using different type scalings proportional to the number of responses, yet there are still imperfections. Ideally I would have been working with per capita data, which is still difficult to represent spatially but also more accurate.

This map could be considered bizarre for a number of reasons. First of all, it is interesting to visualize patterns in the semantics of the health care discussion as you move across the country. Certain words appear in almost every state, as well as among groups who voted both for and against the bill. It is also interesting to pick out what words seem to get the most weight based on how often they appear in the response text. One can begin to pick out trends in the way in which the health care debate is considered and discussed.