As I
did two years ago, I tested that theory this election. I had 11 people send me a picture of their ballot envelope. I then wrote their name in my handwriting. Each voter than copied my version of their signature onto their ballot return envelope. They sent me a picture to ensure it wasn’t their normal handwriting. This simulated signing someone else’s ballot. It’s also legal because each voter signed his or her own ballot.
If signature verification worked, all 11 of those ballots should have been set aside for mismatched signatures. Instead, six were accepted. That’s a greater than 50 percent chance of being accepted. When I did this experiment in 2020 with nine voters,
eight had their ballots accepted. That was an almost 90 percent acceptance rate.
These sample sizes are too small to say definitively that things have gotten better. Let’s hope they have. But either way, one thing is obvious: Signature verification isn’t the fail-safe security measure election officials claim it is.