California Voter’s Choice Act: Utilizing Bayesian Improved Surname Geocoding (BISG) on the 2025 Special Election Voter File

By the UCLA Voting Rights Project in partnership with the California Secretary of State Office

In response to the mid-cycle redistricting proposed in Texas in June 2025, California Governor Gavin Newsom recommended that California redraw its congressional district maps to favor Democrats. As a result, California Proposition 50, officially known as the Election Rigging Response Act, was put on the ballot by the California State Legislature. A special election was scheduled for November 4, 2025, solely for voters to decide whether California should redistrict. Approved by 64.4% of voters, the proposition is an amendment to the California State Constitution that redraws the state’s congressional districts, replacing those drawn by the bipartisan California Citizens Redistricting Commission during the earlier 2020 redistricting cycle. The new districts will be utilized for elections starting in 2026 and continuing through 2030. After the 2030 census, the authority to redraw congressional districts will revert to the independent commission, as part of the standard decennial process.

This report provides a comprehensive analysis of voter registration and turnout across California and in the Voter’s Choice Act (VCA) counties during the November 2025 Special Election. Signed into law in 2016, the California Voter’s Choice Act (VCA) ushered in an election model that expanded early voting opportunities and ballot return methods. The purpose of this analysis is to report patterns by voter characteristics for all California voters. We evaluated the statewide voter file by race/ethnicity, age, language, nativity, voting method, and status as American Indian. The data and tables here are meant to be instructive about how Californians vote and to assess any differences between statewide averages and VCA counties.

In the 60 years since the passage of the Civil Rights Act and Voting Rights Act published statistics have documented a voter turnout gap between White (non-Hispanics) and Black, Latino, and Asian American voters. However, outside former jurisdictions in the Southern U.S. subject to the special provisions of the Voting Rights Act of 1965, almost no other state in the nation has attempted to collect and tabulate data on voters’ race or ethnicity to learn more about voter participation patterns. The efforts by the California Secretary of State, consistent with the VCA model, represent the single largest effort of any non-Section 5 state to document and understand racial and ethnic voting patterns. With more than 22 million registered voters and the most diverse population of any state, California is a critical state for policymakers to understand how voting differs across counties and across diverse communities.

Our analysis categorizes race and ethnicity into five distinct groups: White, Latino, Black, Asian, and ‘Other.’ These categories align with those used by the U.S. Census Bureau and reflect the self-reported classifications utilized in California’s voter files. American Indian and multiracial individuals are included under ‘Other.’ However, we employ geographic analysis methods to gain deeper insights into the American Indian electorate.

Throughout this report we examine the differences in voter turnout among these racial and ethnic groups, commonly referred to as ‘the racial turnout gap’ by political science scholars. This turnout gap is a well-documented issue in voting and elections, both nationwide and in California. Due to historical exclusion and discrimination, along with significant socioeconomic disparities and lower rates of engagement from candidates, communities of color have historically faced lower rates of voter registration and participation compared to White, non-Hispanic individuals – a trend established over the past 50 years in California. This turnout gap is not unique to California and is not solely a consequence of the state’s current electoral system.

Continue reading here…