Re: Printing Vote-by-Mail ballots at home (was Observing process in Placer County CA 4th CD -- Brown v. McClintock)

From: Arthur Keller <voting_at_kellers_dot_org>
Date: Wed Nov 12 2008 - 07:59:11 CST

I understand that about 50% of all ballots cast in California other
than in Los Angeles County are Vote-by-Mail ballots. In Santa Clara
County, California, the figure is about 70%. In Los Angeles County,
the figure is closer to 5% because they cannot easily handle
Vote-by-Mail, and the LA County RoV has not promoted Vote-by-Mail.

Looking at the figures for Santa Clara County ( ), there were at
least 383K Vote-by-Mail ballots cast (they are still counting them)
and only 260K precinct ballots cast (they are finished counting
those, other than provisionals). While you can do individual scans
of ballots at precincts by feeding ballots into the scanner one at a
time, it is not feasible to hand feed a large number of Vote-by-Mail
ballots in a timely manner.

This indicates that Vote-by-Mail ballots should be on heavy stock, so
they can be scanned reducing the problem of jamming and shredding of
ballots. And yes, I have had ordinary 20# paper shredded in ordinary
paper feeders for copiers as a result of a jam.

Santa Clara County, California allows Vote-by-Mail voters to
surrender their paper ballots and vote an ordinary in-precinct ballot
in their *home* precinct. If they do not have a paper ballot to
surrender, or the voter is not in a home precinct, then the voter
must vote provisionally. A voter with a Vote-by-Mail ballot who lost
the envelope may obtain a replacement envelope at the polling place
(as Vote-by-Mail ballots may be returned to any polling place in the
county on Election Day). Vote-by-Mail envelopes are not considered
numbered controlled documents the same way that Vote-by-Mail ballots
are. (The numbers are on a stub that is torn off prior to tallying,
preferably by the voter.)

Alan's idea of printing Vote-by-Mail ballots using the Internet to
print ballots may be reasonable for "facsimile" ballots for overseas
voters, where the ballots have to be manually scanned after being
verified, and are essentially treated as provisional votes. When
treated as provisional ballots, there is no need to digitally sign
any ballot, as the provisional process will ensure that exactly one
ballot is cast by the voter.

If even 20% of LA County ballots were Vote-by-Mail, we'd need a high
volume approach to scanning them. By the way, Santa Clara County
manually compares signatures on the voter registration form against
the signatures on Vote-by-Mail and provisional envelopes.

Also, I think we need to better understand the use of bar codes to
encode the voter's selections. Doing so makes the voter's selections
opaque to the voter, as the written text is used only for hand
counting not the original count. Instead, if the OCR-A font is used
to print the Summary Printed Ballot, would it be better to scan the
actual text to determine the voter's selections than rely on an
opaque-to-the-voter bar code?

Best regards,

At 4:26 PM -0800 11/11/08, Ronald Crane wrote:
>Ballots are different in almost every conceivable way from boarding
>passes. Ballots must be anonymous; boarding passes must not be.
>Election vendors have significant incentives to falsify your ballot,
>the votes recorded on it, and the way it's counted; airlines have no
>incentive to print false or defective boarding passes, and airport
>officials have little incentive to accept them [1]. Hackers have
>significant incentives to attack both, and ordinary PCs' security is
>so poor that they often will succeed. SSL is basically ineffective
>against attacks that compromise your browser; they can redirect your
>traffic to a hack site and then say that it's "secure" and "signed"
>by someone "traceable" to a "valid" "root certificate". Then there
>are social-engineering attacks. Phishing exists for a reason: it
>works. And it will work just as well to get people to give away
>their voting credentials as it does to get them to give away their
>bank passwords.
>Also, I am not convinced that the "bunch of screwed up ballots"
>issue is a significant problem, nor that home-printed ballots would
>not make the problem worse. Just wait until people print their
>ballots on inkjets during the rainy season, or in color (!), or
>scaled incorrectly, or cut off because they used portrait mode
>instead of landscape or vice versa, or otherwise broken because
>their browser is using an incorrect color profile, or their printer
>is set to "economode", or it can't print close enough to the paper's
>edge, or the paper fed incorrectly, or the drum has a little toner
>leak, or they printed the thing on onionskin paper, or who knows
>The best use -- and maybe the only good use -- for this technology
>is to assist voters who cannot otherwise vote independently, even
>with, e.g., a Vote-PAD.
>[1] And also, apparently, little incentive to check them; Bruce
>Schneier recently faked up one and used it to finagle his way into
>an airport's secure area.
> .
>Alan Dechert wrote:
>>I think it should be possible to do this with reasonable security.
>>I can print from home my boarding pass on Southwest (also uses
>>PDF417 barcode).
>>It differs from "polling-place e-voting" in that the vote is not
>>stored electronically.
>>Alan D.
>>>Alan Dechert wrote:
>>>>...2) If we are going to have absentee voting, the ballot should
>>>>be machine printed like the OVC pollsite system (we need to
>>>>develop a system to transmit a form electronically for the remote
>>>>voter to make selections and print the ballot). Optical scan
>>>>systems are not too bad when employed at the pollsite since they
>>>>can catch anomalies like overvotes and stray marks. But you run a
>>>>significant portion of hand-marked ballots though the postal
>>>>system, you're going to end up with a bunch of screwed up
>>>>ballots. This is not a big problem for contests won by a large
>>>>margin, but when you get under 1/2 of one percent (like our
>>>>congressional race), you have a problem.
>>>I'm not sure I understand this proposal. If it is for voters to
>>>print machine-filled ballots on their home printers, it opens the
>>>door to all of the attacks inherent in polling-place e-voting, and
>>>then adds the internet hazards of viruses, network stack
>>>vulnerabilities, social engineering attacks, and the like.

Arthur M. Keller, Ph.D., 3881 Corina Way, Palo Alto, CA  94303-4507
tel +1(650)424-0202, fax +1(650)424-0424
OVC-discuss mailing list
By sending email to the OVC-discuss  list, you thereby agree to release the content of your posts to the Public Domain--with the exception of copyrighted material quoted according to fair use, including publicly archiving at
= The content of this message, with the exception of any external 
= quotations under fair use, are released to the Public Domain    
Received on Sun Nov 30 23:17:15 2008

This archive was generated by hypermail 2.1.8 : Sun Nov 30 2008 - 23:17:22 CST