Washington State Algorithmic Redistricting

This project applied algorithmic methods to create an optimized redistricting plan for Washington State. The goal was to ensure a fair balance of population distribution across districts while maintaining electoral fairness and minimizing political bias.

Project Goals

• Develop an optimized redistricting plan that adheres to legal boundaries while ensuring balanced populations.
• Minimize human bias by leveraging unsupervised machine learning methods.
• Incorporate racial and population balance across districts.

Challenges and Solutions

Challenge: Balancing the population among districts while accounting for racial representation and geographic constraints.

Solution:

• Used 2020 census data to define population and racial constraints for each district.
• Generated a county adjacency matrix to account for geographical restrictions.
• Applied linear programming to minimize the population disparity between districts while maintaining electoral fairness.

Technologies and Tools Used

Languages/Technologies: Python, SciPy

Libraries: Pandas, Plotly, Matplotlib

Other Tools: US Census Data, Adjacency Matrices

Data Description

Data Source: 2020 US Census Data for Washington State population and demographics.

Features: Total population, racial demographics, and adjacency matrix based on geographical data.

Methodology

Data Preprocessing: We built an adjacency matrix for counties and calculated the population and racial composition for each district. We then applied linear programming to balance population while adhering to adjacency constraints.

Optimal Redistricting Plan: 10 districts were generated, ensuring minimal population disparities while adhering to racial balance constraints.

Key Results

• The optimized plan reduced population disparities between districts.
• Key Districts: King County was divided in a way that minimizes overrepresentation.

Data Visualization

Interactive Map and More Information

Explore the interactive map and additional details by clicking the link below:

Analysis & Interpretation

The algorithm successfully balanced population disparities while adhering to racial demographics. However, certain challenges remain, particularly in sparsely populated areas where geographical constraints limited flexibility.

Conclusion

The project provided a fair and balanced redistricting solution that minimized population disparities and adhered to legal boundaries. Future work could include refining the model by incorporating more granular data, such as census tract information.