Re: Shamos Rebuttal, the Finale

From: Anand Pillai <abpillai_at_gmail_dot_com>
Date: Sun May 15 2005 - 02:14:55 CDT

Hi Ron (and other authors of this paper),

      I was going through the rebuttal document and found this
interesting mention of Indian EVMs at section 4 ("Miscellaneous
Errors") of the document.

"Finally, in his 1.5, Shamos implicitly argues that DREs are suitable
for America because India has chosen to use them nationwide. The
authors fail to understand Shamos s intent here. Perhaps he will issue
an updated version of his paper explaining his argument more clearly?"

     I have not read Shamos' original paper, but there are a few key
differences between the Indian EVMs that were used in the Indian
general elections last time and the American DREs, like the Diebolds.

1. First of all, the Indian EVM is basically a hardware device with very little
software components. All the logic of the device is embedded into
firmware, a microchip specifically designed for the purpose of
recording the votes. Hence the opportunities for tampering with the
voting 'software' is very little or non-existent.

2. It has a ballot unit and a control unit. The control unit has a
'close' button which when pressed will freeze the voting firmware in
the system so that it becomes read-only. This is done in a polling
station at the end of voting, so that the system cannot be tampered in
any way later.

2. The amount of intelligence in the voting system itself is minimal.
The voting system has no idea of the candidates, their names etc. The
system can be considered essentially as a 16-way electronic slotting
system, which can record upto 3,840 votes in all these 16 slots. The
balloting officers put a strip of paper along side each of the 16
slots on the machine (or lesser, if the number of candidates is less
than 16). The voter pushes a button next to the slot with the name of
his candidate which records his/her vote for him. At the end of
polling,
the polling officer presses a button on the control unit, which shows the
total number of votes polled against each 'slot'. It is now the job of
the polling officers to take the machines to a tabulation center,
where the tabulation officers look at the names of candidates against
each slot and manually add up the votes from each polling station.

3. Thus the system, though technically a DRE does not preclude the
human role completely since tabulation of votes is done manually for
each DRE. It is not a fully automated system as an American Diebold.

4. Since each ballot unit can record only upto 3,840 votes, the chances that
a corrupt party can influence voting results by capturing a polling
booth (which is quite common here!) is vastly reduced, when compared
to the old papor balloting system. Once the polling officer has closed
a ballot station, the firmware is effectively untamperable.

5. Thuis the Indian e-voting systems are designed entirely different
when compared to American DREs. They are also much less complex with
no GUIs,
databases and associated software components. Hence chances of someone
actually tampering with the 'software' just before an election is much
less in an Indian election when compared to an American one. The only
way to tamper with an Indian DRE is to introduce malware into it right
at the production stage, when the code for the firmware is written.
This however is very unlikely since there are strict controls in
place. Of course, the second way to influence the voting results is to
influence the tabluation officers themselves so that votes are
incorrectly tabulated for a candidate. But again there are different
(human) layers of checks and cross-checks in place to avoid
malpractices such as these.

Even with such as simple system and effectively tamper-proof DREs,
there are quite a few people here in the academia who are unhappy with
the Indian DREs
and recommend systems with a paper trail.

However my point is that the Indian EVMs and American DREs differ so
much that any comfort that Shamos draws from the fact that India uses
DREs in voting
cannot be used as a direct argument for promoting the use of the same
in American elections due to the vast technical and political
differences between the two systems.

Here are few links where you can find more information on the Indian DRE.

About the Indian EVM:
http://www.cs.princeton.edu/~sudhakar/weblog/cache/2004/april/21/06espec.html

About the non-use of paper trails in Indian EVMs

http://economictimes.indiatimes.com/articleshow/654834.cms
http://frazer.rice.edu/epit/documents/dan/Dan-Comment-EK4.doc

Regards

-Anand

On 5/14/05, Ron Crane <voting@lastland.net> wrote:
> This is the (really) final version, signed by all authors. Feel free to
> use it wherever appropriate. We will submit it for publication in
> appropriate peer-reviewed fora; suggestions on where are welcome.
>
> -R
>
>
>
>
> P.S. Note to other authors: this version corrects "George V. Bush"
> (where did *that* come from?) to "George W. Bush," so please pick up
> this version for publication on your websites.
>
> _______________________________________________
> OVC discuss mailing lists
> Send requests to subscribe or unsubscribe to arthur@openvotingconsortium.org
>
>
>

-- 
Anand B Pillai,
Technical Specialist,
SETLabs, INFOSYS Technologies,
Electronics City,
Hosur Road, Bangalore - 560100.
_______________________________________________
OVC discuss mailing lists
Send requests to subscribe or unsubscribe to arthur@openvotingconsortium.org
==================================================================
= The content of this message, with the exception of any external 
= quotations under fair use, are released to the Public Domain    
==================================================================
Received on Tue May 31 23:17:36 2005

This archive was generated by hypermail 2.1.8 : Tue May 31 2005 - 23:17:52 CDT