Editor’s note: Votebeat analyzed state and local voting data, using U.S. census data for context, to examine which voting blocs and counties drove the record ballots cast in the general election. Here is how we did the analysis.

Statewide data on ballots cast, registered voters and voter turnout were obtained from the California Secretary of State’s website on Dec. 6. At that time, no counties reported any outstanding unprocessed ballots. 

Voter registration figures were based on the Secretary of State’s latest registration report, which is as of Oct. 19, the “traditional” deadline to register to vote in the state before the election. Voters can still complete “same day” registration after that date.

For its analysis of voting participation by demographic patterns, Votebeat obtained data from three sources: county election offices, the state’s redistricting database and the U.S. Census Bureau.

Votebeat obtained precinct boundaries and precinct-level voting and registration data for the 2020 election from the 12 most populous counties in California. San Diego County registration and voting data may be slightly lower than final totals because it is based on votes in the presidential election. (A small number of voters likely skipped the Trump/Biden race altogether and only voted in down ballot races). 

Precinct boundaries and precinct-level voting and registration data for the 2016 election came from the state’s redistricting database. Registration totals listed in the redistricting database for counties often differed slightly from numbers reported by the California Secretary of State.

U.S. Census Bureau data on poverty, education and race/ethnicity came from the 2014-18 American Community Survey, at the census tract level.

There are about 8,000 census tracts in California. Votebeat classified the 2,000 census tracts with the highest proportion of residents in certain demographic groups into certain categories. For instance, the 2,000 census tracts with the highest poverty rates in the state were labeled “high poverty areas.” Likewise, the 2,000 census tracts with the lowest poverty rates were labeled “low poverty areas.”

Voting precincts and census tracts often don’t exactly match — they cross boundaries.

Votebeat used mapping software to determine the center point of each voting precinct in the state’s 12 largest counties. It then placed those points within their corresponding census tracts, classifying them as high poverty, low poverty, high proportion of Californians of color, low proportion of Californians of color, etc.

Precinct boundaries can change between elections. Likewise, so can the demographic makeup of a community. At the county level, some of the shift in voting patterns by demographic groups may be due to shifts in demographics or precinct boundaries between 2016 and 2020.

This coverage is made possible through Votebeat, a nonpartisan reporting project covering local election integrity and voting access. In California, CalMatters is hosting the collaboration with the Fresno Bee, the Long Beach Post, and the UC Berkeley Graduate School of Journalism.

We want to hear from you

Want to submit a guest commentary or reaction to an article we wrote? You can find our submission guidelines here. Please contact CalMatters with any commentary questions: commentary@calmatters.org

Phillip Reese

Phillip Reese is a data journalist, an assistant professor of journalism at Sacramento State and a reporter for The Sacramento Bee.