ŷ���ڱ���Ƶ

 

 

 

 

Gerrymandering refers to drawing legislative districts so that one political party wins a disproportionate number of seats.  But how can we quantify it?  One promising approach uses computational and statistical tools to compare a districting plan to an “ensemble” of many possible plans.  Precinct-level election results are overlaid on each plan in the ensemble to determine how many seats each party would have won under that plan, and the results are plotted on a bell curve.  If a particular plan is an “extreme outlier” on this bell curve, this is considered strong evidence that the plan may have been intentionally gerrymandered to achieve a specific partisan outcome.

 

This analytic approach can account for a state’s particular political geography more effectively than simpler measures, and it has already been used in several major court cases regarding redistricting.  The challenge is that this analysis requires lots of data about small geographic units (e.g., voting precincts): geographic information, population data, and election results for each unit.  This data is surprisingly hard to collect; much of it is available only at the county level, and compiling the data for the entire state can be a laborious process.  

 

Fortunately, it turns out that we don’t have to start from scratch!  During the summer of 2019, a student research group at ŷ���ڱ���Ƶ College collected and processed precinct and election data for ŷ���ڱ���Ƶ's 2018 election.  This data is now ready to analyze, and some initial analysis is already in progress.

 

For this project, students will “go back in time” and gather this data for ŷ���ڱ���Ƶ's elections from 2016 and (if time permits) 2014 so that we can extend our analysis to understand recent historical trends.  The main tasks will be:

 

(1) Determine which precinct boundaries have changed between 2010 and 2018, and construct historically accurate precinct maps to match with election data for the 2016 and (if time permits) 2014 ŷ���ڱ���Ƶ elections.

 

(2) Join precinct maps with election and population data to create a clean data set that can be used for analysis.

 

(3) Use analytical tools developed by the Metric Geometry and Gerrymandering Group to analyze how the enacted plans perform vs. an ensemble of districting plans. 

 

The results will contribute to a longer-term effort, joint with Beth Malmskog of ŷ���ڱ���Ƶ College and her students, to understand the current state of ŷ���ڱ���Ƶ’s legislative districts and to provide quantitative information that may help inform the next round of redistricting in 2021.

 

 

Back to the Experimental Mathematics Lab