I'm opposed to electronic voting systems (for critical political use), for the same reason I think relying entirely on electronic copies of books is a bad idea.
As soon as anything exists entirely as electronic digits, the only way to be sure the information has not been tampered with, is to personally have complete end-to-end control of the data, and use strong encryption plus checksumming. Which in itself opens a complexity can of worms.
With electronic voting systems involving thousands of machines and a huge network for collecting and totalizing the results, there's no conceivable way to ensure honesty. Especially when the companies making and programming the machines are known to be partisan, and the groups publishing the results have no legal obligation to tell the truth (as is the case with US media corporations.)
Also, with no permanent physical record of individual votes, there's no possibility of meaningful recounts, or historical archiving of the original votes.
And lastly, without paper ballots there's no way for people to give informal input, such as write-in candidates, rude comments, rants, etc. This is actually a really serious lack, if you want to avoid an eventual violent revolution.
That America's voting system relies so much on electronic systems, should convince anyone with an understanding of how such machines work, that their true reason for existence is to enable vote fraud. Fraud which has been the norm for decades in US elections. Thanks Diebold and their kin.
See
http://blackboxvoting.org/ and
http://everist.org/archives/links/__Vote_fraud_USA_links.txtThere have been so many demonstrations of how the various machines can be easily rigged. Also many documented examples of them actually falsifying votes, both at the individual machine level and vote totalizing level.
One interesting aspect, is the 'vote flipping' issue. In which many people have been reporting (and posting videos of) them touching one screen box, and the machine selecting a different one. goog: 2016 election video vote flipping.
Example:
http://everist.org/archives/links/vote_flip.webm (webm uploads not allowed here.)
There are lots of similar around, I only saved one. But they don't stay online long, for some mysterious reason, ho ho.
The amusing thing about this, is every time that issue comes up in forums there are any number of
sh posters insisting it's just some 'calibration error'. Right... which always just happens to switch votes the same way, in multiple districts with different candidates. Also, sorry, that doesn't wash technically. Touch screens are everywhere now; when was the last time you used an autoteller, office printer control screen, or cellphone, in which touch positioning was way off? So here we get two examples for the price of one: vote-rigging built into the machine software (proof electronic systems can't be trusted), AND forum paid shills and their talking point lists. You can bet if anyone posted a video of the flipping going the other way there'd be endless screaming about 'Russian government hacking the election.'
Which seems to be another talking point list item recently.
I've salvaged two junked office machines lately with touch screens. Both machines were filthy, had been out in the weather. Both touch screens worked perfectly.
Anyway, paper ballots, they are the only reliable way. There can still be ways to rig paper ballot elections, but it's much harder.
It could be possible to combine the two domains, paper and electronic, to get advantages of both (and some other benefits too.)
For instance, fill out paper ballot at polling day, and the ballot paper has these features:
- A unique serial number, printed twice.
- A comment box in which you can write or draw anything you like. This personalizes it for you, so you can recognize it later with no chance of fakery.
- You tear off a slip with the serial number and keep it. Put the other part in the ballot box.
(So the ballot slip is anonymous, but you can still find it in a database.)
Ballots then manually counted, with open public oversight.
Totals done manually, then posted (including all local counts and numbers on rolls), for the election result.
THEN the ballot papers are all fed through digitizers, and saved as actual images.
Then all OCR'd for the tick/number boxes. Second count done electronically. Compared to manual count.
If the two counts are not very close, and make a difference, recount manually and identify discrepancy causes.
But most importantly, put the entire database of ballot paper images online, indexed by serial number.
Result: people can check their ballot actually was counted.
That still leaves the problem of 'dead people voting', imaginary residents, and other ways faked ballots get into the system.
Another negative aspect of electronic voting machines, is the ridiculous cost and waste of materials, to design and build all the machines. That are only going to get used a very few times. Paper is much more resource efficient. Even that scan/OCR stage I suggested would be OK, since it could be done in bulk with just a few high speed machines.