Agreed, if UOCAVA specifically eliminates the anonymity aspect until after reception, we should note that in the functional requirements (I would actually like to show them and indicate they are waived because they are doing that by choice, not be requirement; if someone were to propose a deployment that facilitated doing it WITH anonymity, they SHOULD be open to it).
From: David RR Webber (XML) [mailto:david@drrw.info]
Sent: 2010-07-22 9:11 AM
To: Zelechoski, Peter
Cc: John Borras; EML TC
Subject: RE: [election-services] OASIS NIST UOCAVA Submission.doc
I made some comments to your original on the train this morning - now will add further thoughts.
Notice that UOCAVA voting is DIFFERENT from regular balloting.
The election officials DO KNOW who has cast the ballot. In the existing paper-based process there is the concept of a
double mailing envelope. The outer envelope identifies the voter and their entitlement - the inner envelope contains the actual ballot.
So the process is - election board receives the mail - opens outer envelope - verifies voter. Then pass inner envelope UNOPENED to the staff who count ballots. Now of course it could the same person in reality - but at least the concept allows for some potential voter privacy. It depends on the district / election board of course. E.g. if you are using the generic FWAB form - where your address and everything is a page printed with ballot - its all in the same envelope.
So its important to remember this when looking at electronic interactions. I noticed several of your comments were applicable to polling place style ballots - rather than remote UOCAVA where the election official ARE going to know who you are when you send in the ballot return. Only when the ballot is submitted for counting does it become anonymous again.
This definitely effects how an electronic process can be configured to mimic these interaction patterns.
Added a couple of additional comments I thought of after my initial send.