American election results at the precinct level.

Area of Science:

• Political Science
• Data Science
• American Politics

Background:

• Election results are often fragmented and lack standardization across states.
• Precinct-level data is crucial for detailed analysis of electoral behavior and outcomes.
• Existing datasets lack comprehensive coverage and consistent quality assurance.

Purpose of the Study:

• To create a standardized, high-quality dataset of precinct-level election results.
• To cover nearly all available races from 2016, 2018, and 2020 U.S. elections.
• To ensure the dataset's accuracy by comparing aggregated results to official totals.

Main Methods:

• Collected and cleaned precinct-level election data from all available states and races.
• Standardized data formats and election terminology across different sources.
• Aggregated precinct data to official geographic levels for validation.

Main Results:

• Developed a dataset encompassing almost all U.S. precinct-level election results for three major election cycles.
• Included results for presidential, congressional, gubernatorial, state legislative, judicial, and local races.
• Validated aggregated data against official totals, showing a maximum difference of only 0.457%.

Conclusions:

• The created dataset offers a reliable and comprehensive resource for electoral research.
• The data's high accuracy facilitates in-depth analysis of American elections at the smallest administrative unit.
• This standardized dataset addresses a critical need for consistent, granular election information.