|
|
| Author |
Message |
bru
|
|
A new way to vote.
| 
Nov 11 21:16 UTC 2000 |
Okay...
If visa can keep my credit straight, and so can the other 5000 credit card
issuers in this country...
If Wallstreet can track every last penny I made on my invest ment adn get
it
to me and the other 100,000,000 people invested in it...
Why can't teh federal government get our voting straight.
I propose...
Each state set up a voting card system. Each voter would be issued a voter
card with a specific voter card number on it.
Each card would have storage for enough information so that the voter could
vote at any of a number of different places that allow the input of data into
the card...
That said input could be done at any time prior to the election day...
That said data could be changed at the voters whim...
That said input could be done at any time prior to the election day...
That said data could be changed at the voters whim...
swiped thru a card reader to register the voter, then the card inserted into
a downloader to input the voters choices...
This would speed things up quite a bit. WE could even have drive thru
voting...
You could vote from anywhere in the country...
Following the vote, and removal of information from stolen cards, the states
could then transfer the information to the federal government...
A ballot could be issued like a reciept, signed by the voter, and he could
retain one copy, the other stored for a paper record in case of the need for
a recount.
.
|
| 33 responses total. |
other
|
|
response 1 of 33:
|
Nov 11 21:30 UTC 2000 |
Guarantees of privacy and the secrecy of the ballot (the PRIMARY tenet of our
electoral system) would be absolutely meaningless under such a system.
|
gelinas
|
|
response 2 of 33:
|
Nov 11 21:31 UTC 2000 |
As you note, "they" can do a lot of tracking. This would allow "them"
to personally identify me with my vote. It's a non-starter. As you
pointed out with your throw-away comment on "stolen cards."
Michigan Student Assembly started using web-based voting for its elections
several years ago. The folks developing the application were across the
hall from me, so I heard a little (not a lot, 'cause I didn't much care)
about the various problems they had, including vote fraud. I think that
they got the bugs worked out, but I still don't think it's ready for
anything really important.
Add to this all that we hear about "privacy" and "cookies" and "market
tracking", and people are *still* _seriously_ considering voting this way!?
Are they nuts!?
|
gelinas
|
|
response 3 of 33:
|
Nov 11 21:31 UTC 2000 |
Eric and I are on the same wavelength, I see.
|
aaron
|
|
response 4 of 33:
|
Nov 11 21:45 UTC 2000 |
Why don't we tattoo bar codes on everybody, and then they can simply
scan themselves at any voting center, and cast their ballots?
|
happyboy
|
|
response 5 of 33:
|
Nov 11 21:55 UTC 2000 |
coo-ell!
|
scott
|
|
response 6 of 33:
|
Nov 11 22:54 UTC 2000 |
Kinda weird to hear bru pushing for a national ID card system.
|
bru
|
|
response 7 of 33:
|
Nov 12 00:10 UTC 2000 |
Okay, so you have to have some mechanism to hide the voters identity. Is
there a way to do that?
|
remmers
|
|
response 8 of 33:
|
Nov 12 00:18 UTC 2000 |
If you can figure out such a way that doesn't compromise
reliability, maybe I'll incorporate it into Grex's online
voting program. (Grex's vote program has some features
in common with the system you propose -- vote from
anywhere, vote as often as you want -- but relies on
the honor system for confidentiality. Anyone with root
privileges *could* look at how individuals voted, but
they're not supposed to.)
|
mdw
|
|
response 9 of 33:
|
Nov 12 01:15 UTC 2000 |
The fundemental problem that faces any offline electronic information
system is that it's easy to copy electronic information. This applies
to both voting & digital "coins". There have been, nevertheless, a lot
of proposals on how to do both. Most of them rely on various forms of
cryptography, which has hobbled efforts to publicize & share code,
because the US doesn't want the Libyans to have secure voting
technology. There are also various software "patent" issues, because
whenever people talk about money, the greed factor lights up.
At least one of the URLs I posted before went into some of the issues
that surround computerized voting in general, such as the possibility of
fraud within the computer software, the sort of protection one ought to
deploy around protecting such systems from tampering, etc. In some ways
it's easier to deal with money than voting, because it's acceptable, in
a commercial environment, to make a tradeoff between the cost of extra
protection, and the money saved in reduced fraud due to the extra
protection. It's not clear that's acceptable with voting.
|
otter
|
|
response 10 of 33:
|
Nov 12 01:37 UTC 2000 |
My major problem with #0 is that the election process is handled by the
states, not <thank the gods!> by the fed.
|
scg
|
|
response 11 of 33:
|
Nov 12 03:59 UTC 2000 |
The reason anybody with root on Grex can (but of course doesn't) see how
somebody voted is because the Grex vote program has to preserve that data in
order to allow people to change their votes. The secret ballot systems used
in the US election systems I'm familiar with don't allow people to change
their votes, and thus don't preserve data on whose ballot is whose. An
electronic voting system should presumably keep a list of who has already
voted, and keep a count of the votes, but store them separately. That
preserves secrecy. However, it gets rid of any sort of audit trail, while
the various paper balloting systems in use do keep non-voter-tied audit
trails, so that the votes can be recounted or otherwise examined later.
How to keep the audit trail is an interesting question. Having a paper
printout recording every ballot would be one way to handle it, but if the
paper is printed right when the vote is cast, the ballot list would go in the
order people voted and would be pretty easy to trace. Actually, this may be
a problem with current systems, if the ballots are stacked up in order. I'm
not sure how to handle that one.
Another problem, as others have mentioned, is that voting is handled by the
states, rather than the Federal Government. No problem, let's just let each
state set up one of these systems, and people could vote anywhere in their
state. But wait, there's more. In many areas, such as Florida apparrently,
and California, elections are handled by counties. At that point, you're down
to the level where the daily movements of lots of people have them spending
time in more than one county. My big use for being able to vote anywhere,
instead of at the school two blocks from home where I have to go now, is that
it would be nice to be able to vote somewhere near where I work. But since
I live in Alameda County, and work in San Francisco County, such a system
on a county level would be of rather limited utility to me.
|
nephi
|
|
response 12 of 33:
|
Nov 12 11:19 UTC 2000 |
I was suprised by the awful user-interface of the voting process.
I received no instructions about where to vote. I eventually found the
information from a website that had an invalid domain name.
I waited in this long line at the polling place and witnessed several
people being told that they had to vote somewhere else.
I received no instructions with my punch card. There were none posted
outside of the voting booth, and there were very minimal instructions
inside the booth.
I could barely read much of the text on the ballot, and don't remember
seeing the ballot in any of the papers in read. I had a hard time
finding the non-presedential candidates for whom I wanted to vote.
I felt rushed the whole time due to the long line of people waiting
behind me, and, being in Cook County, I even had to contend with the
butterfly ballot!
While I'm rather sure I got the presidential candidate correct, I bet
that I made at least one mistake while voting. I don't know if I was
victim of any "hanging chad" or "pregnant chad" issues. I didn't even
know to be aware of it until it was discussed in Florida.
I understand why some little old ladies in south Florida had trouble
picking their candidate, and I understand why so many people don't vote,
given this dreadful user interface.
At the very least, if security is a concern, can't we get some nice
little computer kiosks with big lettering and touch screens that we can
use -- even if all they do is generate a printout of our vote that could
be counted later? If those responsible were really clever, the printout
would be readable by humans and computers, so one could verify that she
voted correctly before casting the vote into the ballot box.
I know we can do better than the system we have now.
|
fungster
|
|
response 13 of 33:
|
Nov 12 13:49 UTC 2000 |
There's some serious problems with that, though. It's been mentioned
before that a paper record is necessary, because untrustworthy
software could stuff ballot boxes with no records.
Lest you think this is some paranoid conspiracy ranting
(see http://www.votefraud.org/News/2000/11/110600.html for some
interesting predictions, although they didn't come true),
Douglas Jones, chair of the Iowa State Board of Examiners
for Voting Machines and Electronic Voting Systems (and a
comp sci professor at Iowa) said that it would not be difficult
for one rogue company or programmer to twist the election
their way. It's in Tuesday's RISKS digest.
http://catless.ncl.ac.uk/Risks/21.10.html This fact has been
shown in books such as Votescam, which I used to dismiss as
paranoid conspiracy nonsense (and a lot of it is, but a lot of
the facts that I have checked out are correct, although they the
conclusions they make definitely aren't.) Votescam is a classic
in conspiracy books: http://www.votescam.com to read the first
five chapters, which interestingly focus on South Florida's
election ballot stuffing in the mid 70's.
The scantron (marksense) ballot seems to be a good solution,
although it isn't idiot proof, either. A day or so ago, some Salon
writer posted an article about how she had to deal with a lot of
people who didn't understand how it worked. There, in San
Francisco County, they had machines that took ballots and
immediately checked to make sure that there were no overvotes
or other ambiguities. But, there were three ballot sheets,
due to the myriad initiatives on the ballot, which added to
confusion. Add the people who don't speak English, overzealous
poll watchers, long lines at the booth, and people showing up
at the wrong polling place, and it's amazing how these poll
workers pull it off. Too bad the Salon site is down right now
so I can't pull the article.
As for the most honest system, I would have to agree with the
Votescam book and say that paper ballots, tallied immediately
after poll closing by poll workers with results posted at the
door as soon as voting is counted, then with ballots driven to
the county office and a certain percentage counted again to
make sure that posted amounts are within reason, is the best,
most fraud proof way (at least to prevent computer fraud, although
I think that old fashioned absentee fraud and invalid voter
fraud is still more common). But if you have computers, I agree
with the professor that using an open source dumb terminal system,
perhaps with a cash register receipt showing a token number and paper
version of the choices made to be inserted into a bin like
today, and off the shelf materials would be the fastest, easiest,
and most reliable way. I would use a 386 or 486 running FreeBSD
or Linux with choices presented in text form on the screen
and the user either clicking or entering a letter. Then, after
a user votes for all values, the voter can see what they voted for,
confirm the choices, and then gets a nice little receipt that
they drop in the box for verification. Something like this is
only $300 or so per terminal, with results stored on a Zip
disk or some such on each computer and the disks transferred to
a central site (I don't trust electronic transfer, too easy
to hack) where they are then combined and the results printed out.
Less cost than those new fangled ATM like machines; reuses old
technology; open source to prevent computer vote fraud; and
paper trail in form of receipts to conduct recounts. Sounds like
a winner to me. Why isn't anyone trying it?
|
scg
|
|
response 14 of 33:
|
Nov 12 21:17 UTC 2000 |
I'll agree that a computer voting system would have to be open source. I
think much of the opposition comes from the idea that computer programs are
some dark secret with nobody really knowing what they do, so the programmers
could make them do whatever they wanted and nobody would know. The frequent
problems with Microsoft's closed source software probably do a lot to fuel
this perception. On the other hand, if you make them simple and open source,
with a good audit trail that prints out not only a bar code but also a textual
description of how people voted, that can be examined by the voter or hand
counters, it shouldn't be much harder to figure out whether it's working
correctly than it is with mechanical machines. Having Microsoft create
Microsoft Vote would not be a good idea, though.
I didn't see the ballots they were using in San Francisco. Your description
sounds somewhat like what Ann Arbor uses, but Ann Arbor has much shorter
and less confusing lists of things that have to be voted on.
Having only previously voted with Ann Arbor's system, where voters draw lines
to connect arrows and then feed it into a machine that validates and tabulates
it, I was confused by Alameda County's punch card system. I got handed the
punch card, and asked the poll worker what I should do with it, since the
voting machine seemed very short on instructions. She then stuck it into the
machine for me. Punching the holes seemed pretty straight forward, although
there wasn't any feedback on whether the punching had been successful or not.
I was more confused when I finished, since I was used to Ann Arbor where the
poll workers want the voters to feed the ballots into the machine, and don't
want to see the ballots themselves. In this case, I asked the poll worker
what to do with the completed punch card, and she told me to give it to her.
I didn't quite see what she did with it, but I think she stuck it in a folder.
It didn't strike me as a terribly secure system.
As fungster alluded to, California voting was also generally confusing due
to the length of the ballot. In addition to the usual boards and commissions,
there were something like 30 ballot proposals I had to vote on, for the state,
Alameda County, the City of Berkeley, the Berkeley Unified School District,
and maybe a few othrs. My initial plan had been to vote no on anything I
hadn't been convinced to vote yes on, for purposes of continuing the status
quo, but that strategy failed when I realized that several of these proposals
were to renew things that were expiring, rather than to do something new.
To make matters worse, the ballot language on some of them was rather
confusing, such as the Berkeley Unified School District proposal to "provide
school safety," or the proposal to "correct an error" made in some previous
year. Oh well, welcome to direct democracy.
I also think I may have screwed up my punch card on one of the ballot
proposals. I accidentally stuck the pin in the wrong hole, but had it only
part way in when I noticed I was doing the wrong thing. I then punched the
other hole. I don't think I had it far enough in the first time to have
punched a hole in the card, but since the machine gave no feedback on that
I really have no idea. Anyhow, already being several minutes into my
extremely long ballot, I really didn't want to go back and ask to start all
over again. Hopefully, if it spoiled my ballot, it did so only for that one
proposal. In any case, I don't think anything I voted on came out very close,
so I'm probably one of the few people in this discussion who can honestly say
that my vote didn't make much of a difference.
|
pfv
|
|
response 15 of 33:
|
Nov 12 21:44 UTC 2000 |
There was once a bill (state? federal?) to simplify wording
and PREDEFINE "yes" versus "no" - wtf happened to it?
|
scg
|
|
response 16 of 33:
|
Nov 12 21:48 UTC 2000 |
What do you mean by predefining yes and no? Yes and no seem like pretty
simple concepts. Yes, the proposal passes. No, it doesn't, and the legal
status becomes the same as if it had never been proposed.
|
polygon
|
|
response 17 of 33:
|
Nov 12 23:45 UTC 2000 |
Paper ballots are not as simple and reliable as you might think, given
the enormous variety of marks people make to indicate votes or perhaps
just to frustrate and confuse the vote counters.
In every state, there has evolved an extremely detailed and complex set
of rules for interpreting and counting paper ballots. In general, the
election workers who do the counting are not going to be familiar with
all of the subtleties of these rules, so recounts of paper ballot
precincts tend to result in significant changes from the originally
reported totals.
Experience has shown that recounting paper ballot precincts in the
presence of attorneys for both sides and media results in changes which go
against the party that predominated among those counting the ballots. Not
fraud, but perhaps unconscious bias in ambiguous cases.
In 1960, when Kennedy carried Illinois by what seemed at the time to be a
very slender margin, the Republicans asked for a recount in Cook County
(Chicago and suburbs). The Republican-controlled suburbs used paper
ballots, and by the time the recount was called off, Kennedy had gained
over 2,000 votes.
There have probably also been countless examples of recounts of paper
ballots in Democratic-controlled areas that resulted in major gains for
the Republican candidate. This kind of thing goes both ways.
This kind of experience makes me uneasy about paper ballots as a voting
system.
|
janc
|
|
response 18 of 33:
|
Nov 13 00:03 UTC 2000 |
The modification that I'd like to see to the Ann Arbor voting system would
be a privacy booth when you feed your ballot into a machine, see your votes
on a screen and can push an "accept" or "reject" button. Accept and it keeps
the ballot. Reject and it gives it back and you can try again.
|
gelinas
|
|
response 19 of 33:
|
Nov 13 02:56 UTC 2000 |
Re #14: Your initial intent is why some folks try so hard to get their side
of the question answered "no". Can result in some convoluted ballot language.
|
scg
|
|
response 20 of 33:
|
Nov 13 04:12 UTC 2000 |
Can a no vote ever do anything beyond what would have happened if the proposal
hadn't been proposed?
|
gelinas
|
|
response 21 of 33:
|
Nov 13 04:21 UTC 2000 |
Depends how the proposal is worded. So yes, it can. I can't think of any
concrete examples right off hand. I just remember the controversy over the
language when Miriam's Friends were trying to get assisted suicide approved.
I don't remember the language finally approved.
|
i
|
|
response 22 of 33:
|
Nov 13 05:08 UTC 2000 |
The printed-out, dual-readable ballot scheme that nephi proposed back
in #12 sounds the best to me. You could even set up web sites that let
people fill in their ballot on-line, then print out (Acrobat or similar
*real* WYSIWYG) their ballot to take to the polling place. The scanners
there could compare the entire ballot to (an electronic image of) the
official ballot, verify everything (voted for no more than 3 Assistant
Dog Catchers, right?), etc. Stuffing ballots into the hands of the weak
sounds like it's biggest weakness; must-use kiosks at the polling places
mostly get around that but cut out a lot of the convenience.
|
other
|
|
response 23 of 33:
|
Nov 13 06:43 UTC 2000 |
Ok, taking the various ideas proposed so far, and mish-mashing them together,
this is what I get:
A website and/or free software download which, when run, allows users
to print a user- and computer-readable standardized form which can be taken
to the polls (at which non-networked, dedicated function computers can be made
available to run the same program) and scanned into a ballot scanning/counting
machine. The voters would go through the standard application/verification
process now used, and their total numbers would be counted and used as a basis
for verifying the legitimacy of the vote tally.
The form could be printed at the voter's convenience at any time prior to
election day (within whatever period the election admistrator needs to issue
the correct ballot form information). Voters would still have to vote at
specific, predetermined locations, although I suppose there's no reason why
the system couldn't be set up to allow voters to register a preferred polling
place (say near work instead of near home) at least 45 days before the
election.
Any system which allows complete portability of polling for specific voters
is subject to error either through technological limitations or through
subversive action.
|
polygon
|
|
response 24 of 33:
|
Nov 13 17:26 UTC 2000 |
Michigan, which used to have year-round standard time, had two votes on
whether to have Daylight Savings Time. In 1970, a "yes" vote meant "no
DST", and a "no" vote meant "yes DST". There was a lot of confusion, the
result was extremely close, and the significance of the outcome was
unclear.
In 1972, it came up again, and this time, "yes" meant DST, and "no" meant
no DST. Daylight time won convincingly (though the west side of the state
voted "no"), and we have had it ever since.
|