Signatures. Some are illegible scrawls, and others works of art. Some change over time, and others change every time. No matter their style, or lack thereof, they endure in an increasingly digital age as a legally binding mark, sealing the deal of contracts and proposals. But unbeknownst to many voters, their signature on the mail-in ballot envelope is a critical part of the voting process.      

In many states, voters’ signatures on mail-in ballot envelopes must be verified by election officials against those kept on record. Though this process works in the vast majority of cases, sometimes it can go awry—causing ballots from properly registered voters to be rejected. When this happens, voters in some states—including California—are notified and given a chance to correct the error. Still, the signature-matching process has resulted in thousands of ballots being rejected in California and across the U.S.

The March 2020 primary election in California highlighted the problem. The number of rejected mail-in ballots in that election surpassed 100,000—about 1.5 percent of the votes cast—setting a new high for the state. While most were rejected for lateness, tens of thousands were rejected because of signature issues, drawing attention to disparities in the signature verification processes and inconsistent standards used by counties across California. 

A new law signed by Governor Gavin Newsom in late September aims to address these disparities and reduce the number of rejections of valid mail-in ballots by setting uniform standards across California’s 58 counties for comparing signatures so that every valid vote counts. It also aims to increase transparency by implementing better reporting standards and improving communication between county election offices and voters when a signature matching issue arises.

Critical to the bill becoming law was the sponsorship and support of California State Senator Josh Becker, JD/MBA ’99, D-Peninsula, who recognized that its wide-ranging scope could benefit everyone in the state—including those who are often left behind.

“Challenged signatures disproportionately impact California voters with limited English proficiency, voters living with disabilities, first-time and aging voters, and voters of color,” said Becker. “Senate Bill 503 is about building on what we know works to ensure transparency and fairness.”

What’s more, Becker credits the work of students in a Stanford Law School policy lab for inspiring and informing the legislation.

When Tom Westphal, JD/MA ’21, and four classmates started the Election Law Project in fall of 2018, they knew they wanted to connect students with opportunities to engage in the democratic process—but they didn’t know what projects to focus on. So, they did what smart people do: They asked questions and sought input from their fellow students.

“We were kind of surprised that what came out was that a lot of students were interested in signature verification and vote-by-mail issues writ large,” says Westphal, a U.S. Army officer who served eight years on active duty and now works in intergovernmental relations for the city of San Jose. “It definitely wasn’t something that a lot of voting rights groups or outside experts were talking about at the time.”

Recognizing the magnitude of the task, Westphal and Will Janover, JD ’21, brought the idea to Luci Herman, program director for the Law and Policy Lab at SLS, and Nate Persily, JD ’98, James B. McClatchy Professor of Law and a leading election law expert, who agreed to co-teach a new policy lab with the aim of assessing and documenting signature verification processes in California counties.

In the fall of 2019, the SLS policy lab Every Vote Counts: Voting Verification Project brought together an interdisciplinary group of 15 Stanford students—from the law school and other graduate schools—to interview election officials from 33 California counties, along with several nationwide election administration and voting rights experts, in an effort to understand current county practices for signature verification of vote-by-mail ballots. How do county election officials verify signatures? What steps are taken to allow voters to rectify mismatches? These and other questions uncovered a surprising amount of variety from county to county on everything from the signature envelope design and training for election workers who review signatures to the use of automated technology and how voters can rectify rejected signatures. Whether a voter signature was rejected, they discovered, and the ease with which a voter could remedy a rejection depended on where a voter lived.

“Our research revealed a significant variety of procedures and standards used among the various counties in California in implementing existing legislation regarding signature verification. It became clear that the legislation left substantial discretion to the counties—and that discretion was being exercised in different ways,” says Zahavah Levine, a Stanford Distinguished Career Institute Fellow who left her position as vice president at Google to focus on her concerns about eroding democratic norms in the United States. A member of the policy lab, Levine, who earned her JD from Berkeley and also served as GC of YouTube, later joined Persily’s Healthy Elections Project, a collaboration with MIT, during which she led a study of signature verification across the United States.

In January 2020, the policy lab students published “Signature Verification and Mail Ballots: Guaranteeing Access While Preserving Integrity,” a comprehensive report on the legal and political aspects of signature verification and the processes used by various counties for verifying and fixing non-matching signatures in California. The report included recommendations for California’s county election administrators, secretary of state, legislature, and voters to improve and align signature verification procedures statewide.      

Soon after the report’s publication, Persily and Westphal were asked to join the California Secretary of State’s Working Group on Emergency Election Preparedness, where they submitted recommendations directly to policymakers. Prior to the 2020 presidential election, the California secretary of state’s office adopted those recommendations in emergency regulations that established statewide standards for vote-by-mail signature verification, affecting more than 21 million registered voters.

The new mail-in ballot voter signature verification law adopted by the California legislature in September 2021 builds on the success of the emergency measures as well as the findings in the Stanford report. In large part because of these emergency regulations, the rate of rejected mail-in ballots in the November 2020 general election—in which nearly 87 percent of votes in California were cast by mail—fell dramatically, according to Garrett Jensen (MA/MPP ’20), a policy lab student who helped shepherd the bill through the state legislature as a member of Becker’s policy team.

The new law sets a statewide standard for verifying signatures on mail-in ballots, ensures timely outreach to voters whose ballots are rejected, and requires periodic reporting about the nature of those rejections. “One important point to make is that the law codifies what many California election officials were already doing, but it also creates a standard across the state and more safeguards for voters,” says Westphal.

“It’s one thing to research a topic, identify challenges, and propose solutions. It’s another to actually bring those solutions to life,” says Jensen. “SB 503 would not have happened if not for the efforts of Stanford students of all levels and disciplines. These codified changes will not only improve election administration year over year and cycle after cycle, but they strengthen our democracy.”

The work of the Election Law Project ultimately proved transformative—and prescient.

“In 2019, when we began our research, signature verification was a narrow but important focus. Of course, after the pandemic hit, it proved to be very timely because of the dramatic increase in mail-in voting,” says Persily.

Indeed, the 2020 national election saw a significant boost in mail-in voting—up approximately 46 percent, with one state recording an increase from 4 percent to 40 percent and another from 30 percent to 80 percent; in California, mail voting increased about 50 percent from the 2016 general election, according to Levine.

The policy lab’s findings of significant disparities raised alarms, but their research also provided reassurance. “One of the best parts of working on this was getting to meet so many of the election officials across the state. It was such a source of inspiration that there were so many people working so hard to deliver this service to our voters,” says Westphal. “They contributed to this law too.” SL