LWN: Comments on "Cryptography and elections"

Under pressure from the Australian Government, Professor Teague leaves the University of Melbourne

ras — Tue, 10 Mar 2020 02:19:00 +0000

The speaker who gave the talk being discussed here, Professor Vanessa Teague, has "quit" (the term used by the article linked to below) her position at Melbourne University after the University.

According to her the University came under significant pressure from the government for her role in discovering and making the Health department aware of how Australians can be re-identified from anonymised health care data. The two sources of data are Medicare (Australian's public health insurance system) and the PBS (government subsidised drugs - another part of the public health system). All Australian's will have contributed to the data Professor Teague showed could be de-anonymised on occasion.

Quoting Professor Teague:

> “The uni has to make some kind of decision about trading off the obvious necessity of staying on the right side of the federal government with its funding and its health datasets, versus standing up for [academic freedom],” she said.
>
> They were really stuck in a hard place, which they shouldn’t have been put in, I feel."

https://amp.theguardian.com/australia-news/2020/mar/08/me...

For non-Australians, what Professor Teague is alluding to is a change how Australian Universities are funded. The model has changed from direct government funding via independently overseen grants towards selling their expertise and research capabilities. (In the past Australian Universities have a very poor track record of commercialising their research - I'm not sure what the state of play is now.) The government remains a major funder, however under this new system the arrangement has become one of government departments shopping around and purchasing the research. In this case, the government Health department was purchasing research from the Universities heath researchers and supplying them the data to do the research on.

There may well be more to the story than has been revealed so far, but nonetheless I suspect many who heard Professor Teagues' talk at LCA will be disappointed by this turn of events. For me it was a sharp reminder that while responsible disclosure is almost a daily routine it seems we've got a way to go in bringing even the highy educated liberal elite who run our educational institutions around to that viewpoint.

E-vote advocates still don't get it...

tao — Sun, 01 Mar 2020 09:49:40 +0000

Your reasoning about lack of information doesn't justify opposing postal voting though.

There are 3 scenarios:

1. You only have postal voting
2. You have both postal voting and on-the-day voting (the latter overrides the former)
3. You only have on-the-day voting

If there's new information last minute, 1 is equal for all (just like information 1 day after the election would've been), 2 is status quo with today (people who couldn't vote on the day earlier cannot do that either, so the availability of new information won't affect their vote), 3 is status quo.

Feel free to present actual evidence of cases where there has been election fraud related to postal voting (especially in countries where postal voting is the only--or preferred--option). Postal voting election fraud, at least with the model of postal voting used where I'm from (Sweden) is actually extremely inefficient and we (luckily) don't have voting machines. The most efficient election fraud is probably the Russian model anyway...

E-vote advocates still don't get it...

DonDiego — Sat, 22 Feb 2020 12:35:32 +0000

> But the only reason I'm aware of to oppose postal voting is to let fewer people vote.

No. Postal votes are much easier to tamper with and reduce election security. Also, if people do not vote on the same day, they have different information. I did a postal vote once a week or two before the election and voted differently than I would have on election day because new information surfaced.

E-vote advocates still don't get it...

raven667 — Fri, 14 Feb 2020 18:53:05 +0000

> Uh... What?

Sorry for being obtuse, maybe another way to say it is that even a flawed democratic process results in more public goods as part of the government budget, to "buy" votes, than a non-democratic process. Vote suppression, gerrymandering and other techniques to make the process less democratic hurt the efficiency of translating public opinion into public goods, by reducing the number of voters who need to be appeased. At some point if you continue toward less democracy then you cross over into not-democracy and have less people who need to be appeased with government goods, the end state being authoritarian governments where you really only need the support of the military, police and oligarchs and not the population at large. So the point where "peoples will doesn't matter at all" is much closer to a non-democracy where there are no real elections, and public goods are only handed out enough so that the police/military aren't overworked, than a system where there is just some gerrymandering and you still have to appease 51% the voting public.

Why PR is a problem

Wol — Mon, 10 Feb 2020 01:42:47 +0000

> (b) redrafting the district boundaries invites gerrymandering, which isn't as big an issue in Germany as it is in the US but which would of course influence exactly where the districts are.

Actually, I have no problem with re-drafting boundaries to increase the majority of the incumbent party. One of the important measures of democracy, to my mind, is that people should be represented by the person they voted for, and measures that increase that proportion are okay.

Obviously, there are other criteria that must be met, too, hence I'm in favour of "levelling off" seats, and so on, but the two most important criteria to me are (1) maximising the number of people who are represented by the person they voted for, and (2) maximising the *real* choice of candidates to represent you - that is, a decent choice of people who have a good chance of being elected. The tension between these two is a mark of how democratic the system is, eg "One seat, one candidate" obviously guarantees a perfect mark for the first criteria, but a complete fail for the second. On the other hand, the Italians had a very good PR system that met the second criteria very well, but we all know how stable and long-lived their governments were ...

Cheers,
Wol

Why PR is a problem

anselm — Sat, 08 Feb 2020 02:56:29 +0000

In Germany, half of the 598 seats in the Bundestag are filled by direct elections in 299 electoral districts, and the other half is filled from lists submitted by the states. Every voter has two votes – one to elect a direct candidate in their district, and another to vote for one of the lists submitted by parties in their state. This second vote decides how many seats altogether a party receives in parliament. The three basic cases are:

A party has exactly as many directly-elected candidates as it has seats according to the proportion of second votes it received. Nothing happens.
A party is entitled to more seats in parliament than its number of directly-elected candidates. Any remaining seats are filled from the party's per-state lists.
A party has more directly-elected candidates than it has seats available. Since all the directly-elected candidates have been validly elected, they get to take their seats in parliament. These excess seats are called “overhang” seats. This means that the other parties get additional seats so that the proportions of second votes remains intact, and the total number of seats is increased.

Only parties with either 5% or more of second votes cast or three directly-elected candidates are entitled to seats in parliament in the first place.

For most of the history of the Federal Republic of Germany, there were only three or four parties in the Bundestag, with the SPD (social democrats) and CDU (conservatives) receiving the largest numbers of both first and second votes. Overhang seats were relatively rare. With the rise of the Greens as well as extreme-left (Die Linke) and extreme-right (AfD) parties, there are now more parties which can get considerable numbers of second votes (with CDU and especially SPD losing accordingly) while most directly-elected candidates are still either CDU or SPD, e.g. because people vote for one of the big-party candidates so that their vote isn't “wasted” on someone from a small party who has little chance of being elected directly, but still cast their second vote for a small party, or because a small party doesn't have enough candidates to field one in every electoral district. This leads to larger numbers of overhang seats.

The current Bundestag has 46 overhang seats – 36 for the CDU, 7 for the CSU (the Bavarian arm of the CDU), and 3 for the SPD. There are also 65 other extra “leveling” seats that have been introduced to restore the proportional representation of parties (SPD 19, FDP 15, AfD 11, Greens and Die Linke 10 each). The problem with doing this is that due to inaccuracies in the method used to allocate seats, parties with many overhang seats can gain a slight advantage (which is why the CDU/CSU is reluctant to tweak the system). The leveling seats, which were introduced in 2013, are also necessary to prevent other weird effects that have to do with the fact that list candidates are determined per state; previously, in edge cases, additional votes for a party could actually cause it to lose seats in parliament, and that was ruled unconstitutional by the federal constitutional court.

One obvious method of reducing the size of the Bundestag (which is an issue due to cost, office space, and available room in the debating hall, which was built to accommodate 600 legislators, not up to 800) would be to reduce the total number of electoral districts, but that is not a very popular approach because (a) it means that every representative has a larger swathe of the population to represent, and (b) redrafting the district boundaries invites gerrymandering, which isn't as big an issue in Germany as it is in the US but which would of course influence exactly where the districts are.

Why PR is a problem

ras — Fri, 07 Feb 2020 23:21:36 +0000

You piqued my interest so much I looked up Germany's electoral system up on Wikipedia.

Ensuring proportional representational by appointing new members from under represented parties isn't a solution I've come across before. Every system I've come across before in my limited experience has a fixed number of members in parliament. So there are clearly lots of variations on the MMP theme - more than enough to replace the entertainment provided by heated discussions over FPTP vs IRS.

Germany's system provides in interesting metric - how much they expended the parliament by tells you how many members PR you must elect to make it proportional. Based on the figures you gave (598 directly elected, 709 total meaning 111 had to be added to make it proportional), that's 15%. New Zealand has fixed number of members in parliament, with a fixed 40% being reserved to make it proportional. (The 60% figure I gave earlier with the IIRC proviso just proves I shouldn't trust my memory.) I would have thought 40% was low, but based on Germany's numbers it appears to be plenty.

Regardless of the imperfections you Germans see in your system, externally, from the perspective of someone who lives in a different system, it makes me envious. I guess the proof of the pudding for Germany will be when the current uninterrupted string of centre-left parties are replaced by someone on the right.

That has happened in NZ, and the outcome is what made me take notice of MMP in the first place. They were better than our current right wing government of course - but that would not be hard given they opposed any action on climate change while promoting new coal fire power plants, have handed over young drug mules to Indonesia to be sentenced to death to make a political point (we don't have the death penalty in Australia), opposed cigarette plain packaging laws (which have been a very successful in reducing smoking rates) - it goes on and on. But they were also on a practical level better than any left wing government we've had in the last 20 years. The left wing ones have also made a series gob smackingly dumb mistakes - like _after_ seeing the USA student loan crisis unfold implementing the same thing here. And locally the left mandated that only a licensed a electrician could test an electrical cord with mega ohm meter (this has to be done yearly I think), as a gift to their union financial supporters. It's the sort of thing you can get away with when you can ram anything you want through because you have a single party majority.

After watching both sides pull off stunts like that, you get heartily sick of it after a while.

Why PR is a problem

Jandar — Fri, 07 Feb 2020 22:32:07 +0000

I don't see how this ballooning favors the bigger parties. If the number of seats are proportionally than an increase of 18.56% (111 seats in addition to 598) is 18.56% more seats for ever party. Of course NO current member of the parliament has an incentive to reduce his/her likelihood to grab a seat ;-).

Cryptography and elections

Cyberax — Fri, 07 Feb 2020 20:15:50 +0000

The app didn't help, true. But what it did was highlighting the chaos of the process that in previous elections had just been swept under the rug.

Cryptography and elections

flussence — Fri, 07 Feb 2020 19:42:19 +0000

There was computer tallying of the paper votes - using a questionable proprietary phone app made by an unheard-of company that received money under the table from some of the candidates. The cloud server it phoned home to crashed under moderate load of ~1000 users. And other groups using different software to repeat the same process were getting different results.

All links in the chain need to be secure, or else the votes aren't worth the paper they're written on.

Why PR is a problem

anselm — Fri, 07 Feb 2020 16:22:48 +0000

In Germany the current problem is that there are parties whose number of directly-elected seats is larger than the number of seats they should have according to their proportion of the vote (e.g., the conservative CDU still tends to win quite a number of direct mandates even though their total percentage of votes cast has been dropping). This is counteracted by increasing the number of available seats overall to where the correct relations between parties can be maintained. Theoretically, the Bundestag has 598 seats but the current number of representatives is 709. Apart from ballooning the size of the legislature there are also other less obvious follow-on effects that have to do with Germany's federal structure. In effect, Germany now has the second-largest parliament worldwide (after China).

Urgent election law reform has been called for (and indeed mandated by the German constitutional court) but, perhaps understandably, the bigger parties – who have most to lose if the current practice is abolished – don't fall over themselves getting rid of a system that gives them a considerable advantage.

Why PR is a problem

Wol — Fri, 07 Feb 2020 10:06:25 +0000

> - It's pretty much immune to gerrymandering.

Not quite true ...

But yes this is pretty much the system I was thinking of. That's why I had the party list, sorted by percentage of the vote. That means the politicians can't choose who gets the PR seats - they go to candidates who only lost the constituency by a small margin.

Cheers,
Wol

Why PR is a problem

ras — Thu, 06 Feb 2020 22:26:40 +0000

> look at MMPR. It's used for the ... New Zealand Parliament,

So it's real name "Mixed-member proportional representation". I knew those Kiwi's weren't smart enough to think it up for themselves. Sadly they are smarter than us Ozzie's, because unlike us they use it.

It solves the three main weaknesses of the current system:
- Unlike a single member per electorate, it doesn't devolve into the 2 party system virus.
- It's pretty much immune to gerrymandering.
- Different voting systems (FPTP, IRS, Condorcet, ...) don't have a huge effect.

You might think that last one isn't a problem, but in the Australian State I live in the we have changed out voting system _three_ (count them: 3) times in the last decade or two because the political party notices that a given system (FPTP, IRS) favoured them.

I'm not overly familiar with most of the places that use it, but in the two places I am familiar with (eg, Germany, NZ) the outcome has been nothing short of stellar.

E-vote advocates still don't get it...

milesrout — Thu, 06 Feb 2020 22:03:30 +0000

Postal voting doesn't stop your family members from telling you how to vote and making sure you vote the way they tell you. By far the most important part of the concept of a 'secret ballot' is not just that you don't have to reveal how you voted but that it must be impossible to prove how you voted, because if you can prove how you voted then someone can blackmail or intimidate you into voting for them.

Cryptography and elections

andresfreund — Thu, 06 Feb 2020 21:28:19 +0000

There was no computer voting in Iowa.

Why PR is a problem

farnz — Thu, 06 Feb 2020 20:08:35 +0000

If you want a sense of how you can make PR work well while not breaking the constituency link, look at MMPR. It's used for the Scottish Parliament, Welsh Assembly, London Assembly, New Zealand Parliament, German Bundestag and more.

It works by topping up from party lists when your share of the party vote is lower than the share of constituencies you've won. The idea is that you have one MP per constituency, plus a variable sized "proportional" list that's covering the whole voter base. You can end up in some versions of MMP with a temporarily oversized Parliament, due to "overhang" seats.

E-vote advocates still don't get it...

Cyberax — Thu, 06 Feb 2020 17:05:59 +0000

> But now we are talking about the efficiency of a democratic process at capturing the will of the people and how gerrymandering makes the process inefficient, and not a process where the peoples will doesn't matter at all.
It kinda does.

> Pandering to a gerrymandered subset is somewhat better for the well-being of the general population than only needing the support of the oligarchs, police and military.
Uh... What?

Cryptography and elections

raven667 — Thu, 06 Feb 2020 16:51:43 +0000

> That kind of royal fsck-up can only happen when computers are involved

Yes and no, software commissioned for election events is often of low quality and poorly scoped because work can only start when the money is available during the election and stops once the event happens, a very short timetable, but in this case they had a backup paper process, so we can kind of ignore the software once it was abandoned, and look at how the paper process went and that was it's own mess due to the unnecessary complexity of the caucus process and denial of service attacks carried out by the opposition party on the phone infrastructure used to collect results.

E-vote advocates still don't get it...

raven667 — Thu, 06 Feb 2020 16:43:49 +0000

> panders to some electorally important minority of voters

But now we are talking about the efficiency of a democratic process at capturing the will of the people and how gerrymandering makes the process inefficient, and not a process where the peoples will doesn't matter at all. Pandering to a gerrymandered subset is somewhat better for the well-being of the general population than only needing the support of the oligarchs, police and military.

Why PR is a problem

Wol — Thu, 06 Feb 2020 09:49:13 +0000

> Under PR you need multi-member districts, at least (and proponents of PR will make the districts large so as to "properly" be proportional, so it's very easy to create a situation where poor Dave, who has broadly unpopular views but has also just been violently assaulted by a police officer, finds there's nobody interested in doing anything about it and no recourse whatsoever.

Actually no. My favourite system is one that doesn't, and actually doesn't tinker with FPTP much at all.

Let's make the constituency size 125K, and have one MP per 100K. The House of Commons is about 600MPs for 60M people at the moment ...

Each party submits a list of candidates to Electoral commission. The sum of all votes received by those candidates is the "party vote". After the election, a revised list is created by dropping all except the winners and runners up, who are sorted by percentage of the constituency vote. (This is to get rid of minority parties, sorry :-)

Each candidate is then allocated a PR vote, which is the total party vote divided by their position in the list, and the PR seats are allocated to those candidates with the highest PR vote.

Okay, the Lib Dems would still not have done as well as they should in 1983, but they'd have had about 20% of the seats on a 30% vote.

Cheers,
Wol

A vote without a paper trail can never be trusted

Wol — Thu, 06 Feb 2020 09:25:06 +0000

And in 1983, the Conservative party had one of the largest majorities in the 20th century, with one of the smallest popular votes - smaller even than a vote which devastated the Labour party a few years earlier I believe ...

Cheers,
Wol

E-vote advocates still don't get it...

tzafrir — Wed, 05 Feb 2020 18:06:11 +0000

In Israel the budget of the elections committee is (IIRC) less than ₪100M. This includes all the staff and training, and all the materials. The total cost of the elections to the elections to the state is somewhat closer to ₪500M. This includes the budget of elections committee but also funding various parties.

There are various figures for the total cost of the elections, mainly due to a single day of holiday. Lower estimates extra ₪1,000M and higher ones are ₪4,000M.

So elections here are expensive anyway. Regardless of the method.

E-vote advocates still don't get it...

tzafrir — Wed, 05 Feb 2020 17:50:12 +0000

My 2c from being being more involved in the two recent elections we had.

Elections in Israel are run by the Elections Committee. Headed by a respectable judge, and with people from current parties (there are plenty of those). Various attempts (occasionally successful) for various types of cheating are known to have happened. We had two elections this year and various (potential and practical) methods of cheating have been in the news after the first elections.

Basically the elections are overly expensive (for instance: 4 people are payed be in each at the ballot committee all day. 2 would have been enough). But the point is to trust as few people as possible. Of the committee one is appointed by the central elections committee and 3 are appointed from 3 different parties[1]. So you have 4 people with conflicting interest watching the whole election process.

Counting is a tedious and error-prone process. Therefore a specific procedure was devised on how to count. Therefore all the four people count together and there is no attempt to split the load of the counting between the different members (which would have been faster). There are also various checksums to make stuffing more difficult. But (for instance) if you manages to cooperate with the other party people and manages somehow to get the secretary out of the way, you three may split the result any way you wish.

There are various ways to cheat. Requiring various levels of subversion in the process. But they are difficult. And not leaving traces behind is even more tricky.

I guess that there are many shortcuts one could take to make the process cheaper and faster but less accurate and more hackable (but I guess other countries and jurisdictions have their own methods, sharpened over the years).

I guess this also applies to any electronic voting system: at first shot there will be plenty of loose ends (regardless of the major difference of potentially easier for to make large-scale changes with no proper traces).

----------

[1] And there should be at least one candidate for an opposition party. And attempts are made to keep those people from parties who are not dominant in the local voters population.

E-vote advocates still don't get it...

tzafrir — Wed, 05 Feb 2020 16:41:07 +0000

In Israel elections have always been on a Tuesday. Although the upcoming one will be on a Monday.

Generally the design goal is again to keep things as far away from Shabath (Saturday), on which some of the people avoid doing any kind of work.

Elections as a day off work (hey, it's the Holiday of Democracy. Or whatever). This also makes it possible to use schools for the ballots.

In the recent elections they simply failed to find a Tuesday that was available, and politicians were in a hurry, so they decided to go for a Monday instead.

A vote without a paper trail can never be trusted

jezuch — Wed, 05 Feb 2020 12:22:50 +0000

Hm. A though occurs. How about Asymptotic Representation? The more votes you get the closer you are to 50% of seats, but you can only achieve it with infinite number of votes. All problems solved!!

E-vote advocates still don't get it...

tao — Wed, 05 Feb 2020 09:55:09 +0000

"one folding table costs $30, and each chair costs $10."

You seriously don't have schools, townhalls, or other public meeting spaces that can be used for voting? You know, places that already have chairs and tables? The kind of places where other countries seem to be able to organise their elections just fine without extra costs. Also you might want to check your numbers if $750 blows a budget of up to $2600...

Cryptography and elections

Cyberax — Tue, 04 Feb 2020 22:56:25 +0000

Nope. It actually highlighted a problem that has always existed - the vote totals not matching correctly. It's just that it was the first time these totals were tracked.

So yay for electronic voting uncovering sloppiness (fraud?) that has always existed with paper workflow.

Cryptography and elections

dskoll — Tue, 04 Feb 2020 21:59:19 +0000

Most media reports call the vote a debacle.

It failed in a way paper ballots couldn't have.

Cryptography and elections

Cyberax — Tue, 04 Feb 2020 21:55:40 +0000

What "mess"? The whole kerfuffle is because they would have to wait for results ONE whole freaking day.

This was also compounded by the regular Iowa caucus circus. The attendees elect the leader and then they vote multiple times in person, eliminating one (or more) candidates each time.

So the leader (basically a random Joe, often senior) had to first install the phone app (with bad data connection, it's Iowa) and familiarize themselves with it. Then conduct multiple rounds of voting.

Not a recipe for success.

E-vote advocates still don't get it...

dskoll — Tue, 04 Feb 2020 21:28:04 +0000

You wrote: "The printed paper trail of Nevada's voting systems is more trustworthy than putting pencil to paper."

And then you wrote: "How is it relevant that your system is more trusted than in the US?"

Do please make up your mind as to what you consider relevant.

Cryptography and elections

dskoll — Tue, 04 Feb 2020 21:25:21 +0000

I've avoided replying on this story because it was getting tedious. But these will be my final remarks:

Iowa caucus.

That kind of royal fsck-up can only happen when computers are involved. All the theoretical attacks against paper ballots pale into insignificance with the inevitable mess that will result from computer voting.

E-vote advocates still don't get it...

NYKevin — Mon, 03 Feb 2020 23:24:11 +0000

> We have to win that argument before it makes sense to start arguing about which technology to adopt.

My opinion is that we should fight this on every front simultaneously. I'm uncertain what "winning" that argument would look like from a policy perspective, so I fear that delaying our other tactics would stymie the overall effort. But ultimately, I think we agree on what the world ought to eventually look like and merely disagree on how we might best get there.

> I think traditional methods of voting are basically fine, and the problems we see with them are a result of deliberate sabotage by people who don't like universal voting. Because of that, we should try to fix traditional methods by reversing that sabotage rather than adopting new, untested approaches.

Every state in the union allows some voters to vote by mail. 33 states allow any voter to vote by mail if they feel like it. Seven states (including CA, which is the most populous state) plus D.C. allow any voter to get a ballot mailed to them automatically for every election. Five states (including UT, which is somehow also on the previous list) mail a ballot to every voter, no questions asked. Americans living abroad, including military personnel stationed overseas, generally vote by mail regardless of their state of citizenship (the last state where they lived before leaving the country). There are another 17 states that demand an "excuse" for absentee voting, but frankly, they're in the minority.

Finally, the United States Postal Service (including its predecessors) is literally older than the United States. They have been delivering small pieces of paper across the country for well over two centuries. By now, I would tend to assume that they know what they are doing.

https://www.ncsl.org/research/elections-and-campaigns/abs...

Ranked Choice voting is not a panacea

rgmoore — Mon, 03 Feb 2020 20:35:58 +0000

I have also argued in favor of approval voting. I think it does a good job of being simple enough to understand while also pushing in favor of candidates who appeal to a broad majority of voters rather than a dedicated minority.

E-vote advocates still don't get it...

Cyberax — Mon, 03 Feb 2020 19:59:07 +0000

The problem with "vote buying" is that it panders to some electorally important minority of voters, often putting their interests ahead of public interest.

E-vote advocates still don't get it...

raven667 — Mon, 03 Feb 2020 17:54:55 +0000

> Vote-buying is alive and well even in systems with secret paper ballots. It usually takes the form of campaign promises.

This is what you want in a democracy, legislators "buying" voters support with public policy that benefits the voters, the alternative is legislators buying support from the police or military to stay in power and that's not as much fun.

Ranked Choice voting is not a panacea

epa — Mon, 03 Feb 2020 17:50:27 +0000

FPTP (or plurality, which I think is a less silly name) is at least monotone. That puts it ahead of many ranked-choice alternatives.

I think that approval voting (put an X next to as many names as you like; whoever gets the most votes wins) is a nice simple improvement on plurality voting and keeps monotonicity.

E-vote advocates still don't get it...

NAR — Mon, 03 Feb 2020 14:39:51 +0000

"casting an invalid vote (which is not advertised as an option and a lot of people don't know it exists, and there's a lot of misinformation flying around about it)"

Well, a couple of years ago we had a referendum. The xenophobic agenda was obviously popular, so from the start it was clear which option would win. However, the required turnout was far from guaranteed, so even voting against the government-sponsored option could have helped their cause. That's why there was an actual grass-root, alternative campaign to cast invalid votes. In the end the invalid votes outnumbered the losing option by about 4 times. The ability to cast an invalid vote is important.

E-vote advocates still don't get it...

jezuch — Mon, 03 Feb 2020 13:27:52 +0000

I've long wished the elections had an option like the Debian elections do: None Of The Above. Without it people who feel there's no good choice have to resort to picking the lesser evil, casting an invalid vote (which is not advertised as an option and a lot of people don't know it exists, and there's a lot of misinformation flying around about it), or just not voting (and politicians love that, despite it reducing their legitimacy, because when sensible people stay home, you know who gets the votes). But adding such an option would be political suicide so noone proposes it. Politicians don't like being told that we the people would rather elect a smart toaster, which at least has "smart" in its name.

Why PR is a problem

Cyberax — Mon, 03 Feb 2020 11:00:23 +0000

> there's nobody interested in doing anything about it and no recourse whatsoever.
You have a very optimistic view of the FPTP system. Right now each House representative is responsible for about 750000 people. This means that they likely will not care about you little petty problems (unless they are accompanied by a hefty bribe^W campaign contribution).

It's even worse for the Senate, where in some states one Senate member is responsible for more than 20 million people.

Conversely, even with PR system parties still help individuals because it's a good PR.

Why PR is a problem

pabs — Mon, 03 Feb 2020 04:16:44 +0000

This comment reminded me of the pol.is system used by Taiwan for making multi-stakeholder decisions. Each decision consists of a framing document and an initial set of statements. Participants identify themselves in various ways, may vote agree/disagree/abstain on each of the statements and may add to the list of statements that participants vote on. At the end of the process, principle components analysis and other statistical analysis is done to identify consensus points, groupings within the data etc. The software is FLOSS, I'd love to see it used more in the community.

https://debconf18.debconf.org/talks/135-q-a-session-with-...
https://pol.is/
https://blog.pol.is/
https://github.com/pol-is/