4 ms·

One thing we could do would be to implement some common-sense election integrity improvements such as those recommended in https://nap.nationalacademies.org/cat

One thing we could do would be to implement some common-sense election integrity improvements such as those recommended in https://nap.nationalacademies.org/catalog/25120/securing-the-vote-protecting-american-democracy https://nap.nationalacademies.org/catalog/25120/securing-the..., or similarly by the advocacy group: https://verifiedvoting.org/annual-report-2021/advocating-for-election-security-best-practices/ https://verifiedvoting.org/annual-report-2021/advocating-for.... These should have bipartisan support. Things like paper votes for every election, random manual recounts to sample error rates and confirm they are low, funding improvements to security of voting infrastructure, etc. According to Verified Voting (https://verifiedvoting.org/verifier/#mode/navigate/map/ppEquip/mapType/normal/year/2022 https://verifiedvoting.org/verifier/#mode/navigate/map/ppEqu...), ~5% of votes are made with un-auditable "direct recording electronic" voting machines, where the votes are held in memory rather than printed to paper. When you actually do pentests, it turns out that these machines are sometimes connected to the internet (even though best-practices are to not do so), and they have been demonstrated to be hackable by someone with a USB drive even when air-gapped. Any such tampering is undetectable if done well. It's simply unacceptable to use such technology for voting IMO, and we should pass laws mandating paper-trails and banning DRE voting. One problem is that much of the worry about election fraud is motivated reasoning ("my preferred candidate didn't win therefore there must have been fraud; let me find evidence of fraud to support my theory"), rather than actual concern based on an objective analysis starting with the facts. I do wonder if we can still make some changes to the system that strictly improve security and require zero trust of the opposition party, without destroying trust in the underlying institutions.

piva00 4y ago
I believe you are focusing on a moving target. Electoral fraud is just the current wedge issue, it will become something else next when a lot of effort and thought is wasted trying to fix a system that is not that broken to be causing such uproar. You waste resources playing a whack-a-mole game against the firehose of bullshit. It's electoral fraud now, something to do with green energy plans next, then whatever welfare issue is hot in the moment. Something with immigrants and then start repeating.
- rayiner 4y ago
  It's essential to the voting system that different groups who do not trust each other can trust the vote counts. There is a reason even a high-trust, homogenous society like Taiwan does vote counting like this: https://www.youtube.com/watch?v=cqKt-lPfJuw https://www.youtube.com/watch?v=cqKt-lPfJuw Lack of trust in the election system isn't a partisan issue. When I was a young Democrat, the kerfuffle was over how Diebold voting machines delivered Ohio for George W. Bush. In 2018, two thirds of Democrats polled said they believed Russia "tampered with vote tallies in order to get Donald Trump elected President." Programmers should understand this. When a system is opaque and hard to reason about, people won't trust it. We can't afford for our election system to be like that.
  - godelski 4y ago
    > It's essential to the voting system that different groups who do not trust each other can trust the vote counts. This is one of the big reasons I endorse cardinal style voting over ordinal systems. The act of counting the vote is trivial in comparison (singular round, no runoffs, trivially parallelizable). There are other reasons I advocate for it, but this is a major one: transparency. I definitely a agree that creating a highly transparent system is integral to our elections. I think we need substantially more transparency than we have now. But not just with how votes are counted, but also in who has power and how they are making decisions. The left has been discussing quite a bit about how the right is trying to hijack the system. So I think even both sides would be highly in favor of increased transparency. But like I said in my OP, complex problems require complex solutions. The original post was about one such need. I think this is another need under the same umbrella and I think there are many more too.
- theptip 4y ago
  > you are focusing on a moving target. Election security might be a moving target for you, but this is something I have been paying attention to for at least a decade. I think it's important to note that there is an underlying bipartisan issue here that has been known for a long time. > a lot of effort and thought is wasted trying to fix a system that is not that broken This is certainly the blue-tribe talking point right now, but it's an oversimplification, and I encourage you to actually read and engage with the materials I linked. There are security vulnerabilities and auditability failures in the system which should be fixed, and this was a common blue-tribe issue prior to 2020, when it switched to being a red-tribe issue. I don't think these issues are catastrophic failures that invalidate the election results, but they are serious and worth attention. I think one should separate out concrete claims of fraud in 2020 that have since been refuted by recounts (e.g. Antrim County) as examples of the motivated reasoning I referred to, vs. issues like un-auditable DRE machines which are actual problems that need to be fixed. I do understand where you're coming from; the instinctive reaction in a low-trust partisan environment is that giving oxygen to the bad-faith/motivated-reasoning claims of election fraud does more harm than good. However for one example, I think if the red tribe has a good showing tomorrow (i.e. lose some of their incentive to cry fraud), then that would be an ideal time to get some bipartisan legislation in place to try to bleed the issue of salience ahead of the 2024 elections where it could be extremely dangerous. > You waste resources playing a whack-a-mole The point I was making is that there is some set of policy changes that the blue team thought were beneficial in 2016, and perhaps we could all agree to do them. It's not merely whack-a-mole if you leave the world a better place after whacking the first mole. The goal I'm advocating here is not to play zero-sum games to "whack the opponent", it's to actually make objective improvements to the system that in normal times would be uncontroversial bipartisan proposals. If we can execute the political Ju-Jitsu to redirect the current toxic discourse into actual improvements that would be a major win. Sure, following that there will be another wedge issue; that's politics. But I think the important question is whether we make non-zero-sum gains during each issue cycle, and sometimes it is possible to do so.
mistermann 4y ago
> According to Verified Voting (https://verifiedvoting.org/verifier/#mode/navigate/map/ppEqu https://verifiedvoting.org/verifier/#mode/navigate/map/ppEqu...), ~5% of votes are made with un-auditable "direct recording electronic" voting machines, where the votes are held in memory rather than printed to paper. When you actually do pentests, it turns out that these machines are sometimes connected to the internet (even though best-practices are to not do so), and they have been demonstrated to be hackable by someone with a USB drive even when air-gapped. Any such tampering is undetectable if done well. It's simply unacceptable to use such technology for voting IMO, and we should pass laws mandating paper-trails and banning DRE voting. This is quite contrary to the ~~propaganda~~ trustworthy journalism take on the matter: https://www.theguardian.com/us-news/2020/nov/13/us-election-most-secure-history-voter-fraud-false-claims https://www.theguardian.com/us-news/2020/nov/13/us-election-... > One problem is that much of the worry about election fraud is motivated reasoning ("my preferred candidate didn't win therefore there must have been fraud; let me find evidence of fraud to support my theory") Another problem is that much of the lack of concern about election fraud is based on bad epistemology ("my preferred candidate won fair and square [in an epistemically sound, absolute sense] because some people said it was the most secure in history). > ...rather than actual concern based on an objective analysis starting with the facts. In my experience, this (utilizing actual objective analysis) is a good way to collect downvotes and insults. > I do wonder if we can still make some changes to the system that strictly improve security and require zero trust of the opposition party, without destroying trust in the underlying institutions. I wonder if we will ever try, as opposed to simply repeating the same claim enough times so that people believe it to be true. The latter is certainly cheaper, and certainly gets the job done (where the job is persuading the public that their election system is secure in fact, regardless of whether it is actually true). https://en.wikipedia.org/wiki/Illusory_truth_effect https://en.wikipedia.org/wiki/Illusory_truth_effect
- theptip 4y ago
  > This is quite contrary to the ~~propaganda~~ trustworthy journalism take on the matter: I would note that it's possible for both to be true: A) this was the most secure election in US history, B) there are still some meaningful security issues to iron out. For example between 2016 and 2020, DRE machines were being phased out (28.9 -> 11.4%), so in that regard we do seem to be more secure than in the past. And I want to be clear -- I think the detailed content of that Guardian article is broadly fair, even if they do engage in some sloganeering. > The officials who signed the statement said they had no evidence that any voting system had deleted or lost votes, had changed votes, or was in any way compromised. > They said all the states with close results have paper records, which allows for the recounting of each ballot, if necessary, and for “the identification and correction of any mistakes or errors”. I think this is an honest evaluation of the evidence RE: specific claims of voting fraud in 2020. (Though I'm always interested to hear folks' best evidence to the contrary.) Importantly, absence of evidence _is_ evidence of absence (evidence != proof!). If there was vote tampering, it should have been detected by a recount somewhere. But we don't see that.

piva00 4y ago

I believe you are focusing on a moving target. Electoral fraud is just the current wedge issue, it will become something else next when a lot of effort and thought is wasted trying to fix a system that is not that broken to be causing such uproar. You waste resources playing a whack-a-mole game against the firehose of bullshit. It's electoral fraud now, something to do with green energy plans next, then whatever welfare issue is hot in the moment. Something with immigrants and then start repeating.

rayiner 4y ago
It's essential to the voting system that different groups who do not trust each other can trust the vote counts. There is a reason even a high-trust, homogenous society like Taiwan does vote counting like this: https://www.youtube.com/watch?v=cqKt-lPfJuw https://www.youtube.com/watch?v=cqKt-lPfJuw Lack of trust in the election system isn't a partisan issue. When I was a young Democrat, the kerfuffle was over how Diebold voting machines delivered Ohio for George W. Bush. In 2018, two thirds of Democrats polled said they believed Russia "tampered with vote tallies in order to get Donald Trump elected President." Programmers should understand this. When a system is opaque and hard to reason about, people won't trust it. We can't afford for our election system to be like that.
- godelski 4y ago
  > It's essential to the voting system that different groups who do not trust each other can trust the vote counts. This is one of the big reasons I endorse cardinal style voting over ordinal systems. The act of counting the vote is trivial in comparison (singular round, no runoffs, trivially parallelizable). There are other reasons I advocate for it, but this is a major one: transparency. I definitely a agree that creating a highly transparent system is integral to our elections. I think we need substantially more transparency than we have now. But not just with how votes are counted, but also in who has power and how they are making decisions. The left has been discussing quite a bit about how the right is trying to hijack the system. So I think even both sides would be highly in favor of increased transparency. But like I said in my OP, complex problems require complex solutions. The original post was about one such need. I think this is another need under the same umbrella and I think there are many more too.
theptip 4y ago
> you are focusing on a moving target. Election security might be a moving target for you, but this is something I have been paying attention to for at least a decade. I think it's important to note that there is an underlying bipartisan issue here that has been known for a long time. > a lot of effort and thought is wasted trying to fix a system that is not that broken This is certainly the blue-tribe talking point right now, but it's an oversimplification, and I encourage you to actually read and engage with the materials I linked. There are security vulnerabilities and auditability failures in the system which should be fixed, and this was a common blue-tribe issue prior to 2020, when it switched to being a red-tribe issue. I don't think these issues are catastrophic failures that invalidate the election results, but they are serious and worth attention. I think one should separate out concrete claims of fraud in 2020 that have since been refuted by recounts (e.g. Antrim County) as examples of the motivated reasoning I referred to, vs. issues like un-auditable DRE machines which are actual problems that need to be fixed. I do understand where you're coming from; the instinctive reaction in a low-trust partisan environment is that giving oxygen to the bad-faith/motivated-reasoning claims of election fraud does more harm than good. However for one example, I think if the red tribe has a good showing tomorrow (i.e. lose some of their incentive to cry fraud), then that would be an ideal time to get some bipartisan legislation in place to try to bleed the issue of salience ahead of the 2024 elections where it could be extremely dangerous. > You waste resources playing a whack-a-mole The point I was making is that there is some set of policy changes that the blue team thought were beneficial in 2016, and perhaps we could all agree to do them. It's not merely whack-a-mole if you leave the world a better place after whacking the first mole. The goal I'm advocating here is not to play zero-sum games to "whack the opponent", it's to actually make objective improvements to the system that in normal times would be uncontroversial bipartisan proposals. If we can execute the political Ju-Jitsu to redirect the current toxic discourse into actual improvements that would be a major win. Sure, following that there will be another wedge issue; that's politics. But I think the important question is whether we make non-zero-sum gains during each issue cycle, and sometimes it is possible to do so.

mistermann 4y ago

> According to Verified Voting (https://verifiedvoting.org/verifier/#mode/navigate/map/ppEqu https://verifiedvoting.org/verifier/#mode/navigate/map/ppEqu...), ~5% of votes are made with un-auditable "direct recording electronic" voting machines, where the votes are held in memory rather than printed to paper. When you actually do pentests, it turns out that these machines are sometimes connected to the internet (even though best-practices are to not do so), and they have been demonstrated to be hackable by someone with a USB drive even when air-gapped. Any such tampering is undetectable if done well. It's simply unacceptable to use such technology for voting IMO, and we should pass laws mandating paper-trails and banning DRE voting. This is quite contrary to the ~~propaganda~~ trustworthy journalism take on the matter: https://www.theguardian.com/us-news/2020/nov/13/us-election-most-secure-history-voter-fraud-false-claims https://www.theguardian.com/us-news/2020/nov/13/us-election-... > One problem is that much of the worry about election fraud is motivated reasoning ("my preferred candidate didn't win therefore there must have been fraud; let me find evidence of fraud to support my theory") Another problem is that much of the lack of concern about election fraud is based on bad epistemology ("my preferred candidate won fair and square [in an epistemically sound, absolute sense] because some people said it was the most secure in history). > ...rather than actual concern based on an objective analysis starting with the facts. In my experience, this (utilizing actual objective analysis) is a good way to collect downvotes and insults. > I do wonder if we can still make some changes to the system that strictly improve security and require zero trust of the opposition party, without destroying trust in the underlying institutions. I wonder if we will ever try, as opposed to simply repeating the same claim enough times so that people believe it to be true. The latter is certainly cheaper, and certainly gets the job done (where the job is persuading the public that their election system is secure in fact, regardless of whether it is actually true). https://en.wikipedia.org/wiki/Illusory_truth_effect https://en.wikipedia.org/wiki/Illusory_truth_effect

theptip 4y ago
> This is quite contrary to the ~~propaganda~~ trustworthy journalism take on the matter: I would note that it's possible for both to be true: A) this was the most secure election in US history, B) there are still some meaningful security issues to iron out. For example between 2016 and 2020, DRE machines were being phased out (28.9 -> 11.4%), so in that regard we do seem to be more secure than in the past. And I want to be clear -- I think the detailed content of that Guardian article is broadly fair, even if they do engage in some sloganeering. > The officials who signed the statement said they had no evidence that any voting system had deleted or lost votes, had changed votes, or was in any way compromised. > They said all the states with close results have paper records, which allows for the recounting of each ballot, if necessary, and for “the identification and correction of any mistakes or errors”. I think this is an honest evaluation of the evidence RE: specific claims of voting fraud in 2020. (Though I'm always interested to hear folks' best evidence to the contrary.) Importantly, absence of evidence _is_ evidence of absence (evidence != proof!). If there was vote tampering, it should have been detected by a recount somewhere. But we don't see that.