While “no-excuses” vote-by-mail is offered by thirty-six states and five states have moved to 100% vote-by-mail, there still exists a reluctance to move completely to a vote-by-mail process. Some of this reluctance is based on tradition while another is driven by concerns over security and reliability.
In the process to adopt vote-by-mail, there are many aspects to consider to ensure that processes involved produce the highest level of reliability. There is a technology solution to address the issues of security and reliability called Automated Signature Verification (ASV) fully explored in this white paper.
The key factors explored in these pages are as follows:
- Challenges of Signature Verification including signature matching
- Automated Signature Verification (ASV) and where ASV is in use today
- The Science Behind ASV including its elements, analysis and neural networks
- Verifying ASV Results for Vote-By-Mail
The most common method employed to ensure that each vote is legitimate is the use of signature match or review whereby each signature on a submitted ballot is compared with signatures on voter records to determine authenticity. In the majority of elections that use vote-by-mail or absentee ballots, signatures are reviewed manually. This causes problems not only in the time and resources required, but also with the ability to reliably compare signatures due to staff who are not always comprehensively trained and with claims that the verification process can be politically manipulated.
Automated Signature Verification (ASV) provides a solution to this problem. Parascript ASV machine learning-based software takes one or more reference signatures from voter records and compares them with the signature on the ballot to determine authenticity. The process takes less than a second for each ballot. It also produces results that are not only more reliable than results achieved with even well-trained staff, but are consistent and tamper-proof. Even if signatures look different due to changes over time, the software can adapt – looking at key, proven, hard-to-detect characteristics.