The NYT has this story on racial disparities in rejection rates in Washington state, which is an all mail-in balloting state:
Among the thousands of mail-in ballots that were rejected in Washington State during the 2020 election, auditors have found that the votes of Black residents were thrown out four times as often as those of white voters.
The rejections, all of them because of problematic signatures, disqualified one out of every 40 mail-in votes from Black people — a finding that already is causing concern amid the national debate over voter access and secure balloting. Washington, a state with broad experience in mail-in balloting, found that rejection rates were also elevated for Native American, Hispanic, and Asian and Pacific Islander voters.
State officials said there were no signs that ballots cast by Black or other minority voters were knowingly singled out by poll workers, or that any of the ballots were deliberately falsified; the rejections were a result of signatures that were missing or did not match those on file, a possible result, the officials said, of voter inexperience, language problems or other factors….
The findings in Washington State mirror mail-ballot research that has been conducted in other states in recent years, including Georgia and Florida. But they are crucial in a state like Washington, which in 2011 became the second state to adopt all-mail balloting, behind Oregon. Mail-in voting has been an option for all statewide elections since 1991….
Auditors found nearly 29,000 ballots that were rejected for various signature problems in the 2020 general election — either the signature was missing or it did not match what was on file. Looking for signs of bias, the state examined thousands of accepted and rejected signatures more closely, but the auditors largely agreed with the decisions that county election workers made and found nothing in the way the ballots were reviewed that would explain the disparity.
Other groups that had higher rejection rates were men, younger voters and less-experienced voters of all races and ethnicities. The audit also found that rejection rates varied by county, a difference the auditors said could reflect varying degrees of strictness in matching signatures.