Impact of Gerrymandering on US elections

Gerrymandering is the deliberate manipulation of electoral district boundaries to favor a particular political party, undermining the democratic principle of fair representation. This study investigates the extent and impact of gerrymandering in the United States using Pennsylvania's 2012 and 2018 congressional district plans as case studies. Compactness and fairness metricsincluding Polsby-Popper, Schwartzberg, Reock, Efficiency Gap, Partisan Bias, and Mean-Median Difference-were employed to quantitatively assess district geometry and voter distribution. The 2012 plan, drawn under Republican control, exhibited poor compactness and significant bias, indicating strong evidence of partisan gerrymandering. In contrast, the 2018 remedial plan demonstrated more balanced population distribution and improved fairness scores. To further address redistricting bias, this study proposes a computational approach using weighted k-means clustering, incorporating population constraints to generate more compact and demographically balanced districts. Results show that the weighted k-means method significantly improves population uniformity and compactness while reducing wasted votes, though further refinement is needed to minimize partisan bias. Overall, this research highlights the potential of data-driven algorithms in promoting fairer redistricting and strengthening electoral integrity in the United States.