How Data Visualization Can Identify Political Gerrymandering

Gerrymandering, the practice of manipulating the boundaries of an electoral constituency to favor one political party, dates back to 1812 when Massachusetts Governor Elbridge Gerry signed into law a redistricting plan that was designed to benefit his political party.

In fact, the practice of gerrymandering pre-dates Governor Gerry. In 1788, the Federalist Party, led by Patrick Henry, drew the boundaries of Virginia’s Fifth Congressional District in an attempt to prevent James Madison from winning a seat in Congress.

Since then, the practice of gerrymandering has become widespread across many states and counties. A recent article in The Washington Post by Christopher Ingraham draws further attention to the issue by exploring the level of gerrymandering that’s in place across the United States in the 113th Congress.

To calculate his gerrymander scores, Ingraham compared the ratio of the district area to the area of a circle with the same perimeter, following the Polsby-Popper method. Ingraham then inverted these values and multiplied by 100 to create a 0-to-100 index, with the least compact districts receiving the highest scores.

Using this approach, some areas and states such as the eastern shore of North Carolina, received very high scores (state average of 88.33). By comparison, other states such as Nevada (49.35), received relatively low scores.

According to Ingraham’s analysis, the Democrats are underrepresented in the U.S. House of Representatives by about 18 seats. Interestingly, Democrats won in nine of the 10 most gerrymandered districts. However, eight of those 10 districts were drawn by Republicans.

Intrigued by Ingraham’s results, we at TIBCO Spotfire sought to reproduce his findings. Using Congressional district shapefiles from the U.S. Census Bureau, we calculated the area and perimeter of each district using TIBCO Enterprise Runtime R with the R packages sp, rgdal, and geosphere.

The U.S. Census Bureau uses simplified coastal geometry, so irregular coastlines such as Maryland do not unduly exaggerate the perimeters of coastal districts.

For each district, we calculated the ratio of the district’s area to the area of a circle with the same perimeter as the district. We subtracted those numbers from 100 percent to calculate the final gerrymander score.

Our findings were slightly different from Ingraham’s. For instance, in North Carolina’s highly gerrymandered District 1, Ingraham calculated a score of 96.01 while TIBCO Spotfire computed a score of 95.92.

To learn more about the methodology behind our mashup and the results of our findings, click here to view our interactive map.


Next Steps:

  • Try Spotfire and start discovering meaningful insights in your own data.
  • Subscribe to our blog to stay up to date on the latest insights and trends in big data and big data analytics.