Sophie Guo, Math/PoliSci major, Bridget Dou, ECE/CompSci major, Sachet Bangia, Econ/CompSci major, and Christy Vaughn spent ten weeks studying different procedures for drawing congressional boundaries, and quantifying the effects of these procedures on the fairness of actual election results.
There has already been research done with North Carolina districts, described in http://today.duke.edu/2014/10/mathofredistricting. There, Jonathan Mattingly and Christy Vaughn showed that randomly re-drawing district boundaries would have dramatically changed election results. This summer's team extended the analysis to many more states, and found that states with independent election commissions (like Iowa) had statistically fairer results than states with very partisan districting systems (like Maryland).
See Quantifying Gerrymandering, a website developed by Sachet Bangia, for more details about the project.
Gerrymandering work now posted on Arvix: https://arxiv.org/abs/1704.03360
- Political Science
Undergraduates: Sophie Guo, Bridget Dou, and Sachet Bangia
Faculty Lead: Jonathan Mattingly, Professor, Mathematics
Graduate student mentor: Christy Vaughn, Program in Applied and Computational Mathematics, Princeton
Relatively Prime podcast on the project: