I think the best system is one like New York has. You fill out a paper ballot, you feed your ballot into the scanner, and then the paper ballots are collected.
So you have what is essentially two separate sets of the vote. The scanner count is the official vote but you can compare the scanner numbers with the paper ballots. If the two totals don’t match, then you know you have a problem. You may not know whether it was the scanners or the paper ballots that were tampered with but at least you’ve identified that a problem exists.
In order to have an invisible fraud, you’d have to tamper with both the scanners and the paper ballots and make sure that all of your tampering matched.
Wisconsin is now set to recount the votes. So, based on the way Wisconsin votes and the way the recount is being performed, would this be sufficient to detect evidence of fraud there?
Actually, I moved out of California in between the primary election this year and the general election. I didn’t notify the county elections board, but I got a postcard in the mail from the elections board informing me that unless I verified my address in California I would be stricken from the voting rolls.
I suspect placing a forwarding address on my mail triggered this action.
I imagine it depends partially on how the recount is done. Say some voting machines are compromised to record votes incorrectly. If the recount only looks at the paper only ballots or only at the tabulation in the central location, it may miss signs of individual machines being compromised. Another possibility - say the machines that are compromised don’t produce a paper audit trail. Then, the recount might just be recounting fraudulent votes and coming up with the same result. Since we have so many types of voting and different voting authorities, it’s very difficult to say generally if a recount would detect fraud.
The Wisconsin Elections Commissions allows municipalities to select voting systems, which may either be paper ballots or electronic voting systems with a paper audit trail. The 2014 audit results (which you can find here) show a handful of irregularities, primarily due to ballots jamming when fed into the scanner. Pennsylvania has a combination of paper ballots and electronic voting with no paper audit trail. The paper ballots can be recounted and voting machines examined to see if the version of firmware is the same as specified in a pre-election verification (which I assume they did) but unless there is sign of clear tampering at the polling station or tally level I doubt there is much that can be done. Michigan is all paper ballots, and they just completed their official tally right before Thanksgiving with just over a net 10k votes to Trump out of around 4.8M votes. A recount could be conducted and there is an outside possibility that it could flip, but for the vote to be reversed all three states would have to flip, which is statistically unlikely without finding evidence of systemic fraud. This is really a stalling action (and I don’t think it can hold up the Electoral College vote if the paper recount isn’t complete unless there is substantial evidence of bias or fraud), so I wouldn’t be holding out any hopes.
Note that although it is not generally the case that third party candidates can really spoil an election unless they are literally splitting a party vote (rather than taking votes away from both candidates by some total fraction of the third party voters), in this case all three states votes are close enough that the election could have potentially shifted to Clinton. While I don’t think this makes the case that voters should not vote for a third party candidate if they sincerely believe in their vision as an expression of support, ‘protest voting’ against a hated candidate by selecting a third party is ineffectual and even counterproductive under the Electoral College system as demonstrated here. It is unfortunate that the way the Electoral College has developed, along with the polarization of politics in this country, essentially marginalizes anyone who doesn’t significant side with either major political party but that is the system we have, and is something to bear in mind for future elections.
Or if you obtained a driver’s license, since doing so in many states automatically registers you to vote and informs the election commission of your previous state that you have changed residence.
I don’t approach this as someone who expects this to change the 2016 election, but that there was a sufficient amount of hacking in this election that we damn well need to know whether that extended to the election itself to prepare for 2020. In other words, I don’t want a recount with the hope of changing the past results, but to be able to trust the results moving forward.
In that case you should support a move to end-to-end verifiable secure voting systems to replace the profusion of different paper ballot, optical scan, and electronic recording electronic voting machines. A public key encryption system using open source code that can be freely evaluated by any observers, in which the source code is verified at the time of vote by generating a checksum of the code itself plus the ballot would be verifiable from ballot to tally point, and the process itself would be transparent and essentially tamperproof. It wouldn’t require any proprietary equipment other than a standard inexpensive computer (for code this simple you could literally run it on a Raspberry Pi or BeagleBone) and attacking the physical system could not feasibly compromise the integrity of the ballot; at most, it could only cause the ballot to not be correctly submitted, which could be addressed by message receipt verification as part of the process. The system could implement an auditable paper trail by attaching the device to a printer, or even by having the user generate and optical scan record as part of the voting process (e.g. one punches the ballot and inserts it into the device which scans it as it feeds it into a secure hopper) but quite honestly this system is similar to and less complex than the financial transaction system that drives trillions of dollars of monetary transfers a day.
Stranger’s excellent post touched on election rigging methods that probably were not decisive, so I took the liberty of emphasizing his mention of methods that might have been decisive.
Perhaps “red vs blue” is not the best description of our polarization. We’re a nation of the gulled versus the galled.
That’s possible - timing would seem very efficient if that were the case - I think I received the postcard around 4 business days after applying for the new license, which means they would have had to send it almost immediately. On the other hand, it was a couple weeks after the mail forwarding was submitted. Maybe they’re just on the ball.
It’s possible they were just that efficient: Everything other than the sending of the card itself would probably have been done electronically, and hence nearly instantly.
But given the whole voting machine controversy of 2000, with implications that presumably GOP-aligned businesses (Diebold being the most prominent) tampered with the machines they made, wouldn’t there be another source of correlation? Elections are run at the county level, and a county that tends Democratic would also tend to pick paper/optical balloting out of concerns about voting machines, while counties not overly concerned about voting machines would tend GOP.