In 2009 the German Constitutional Court (similar to the U.S. Supreme Court) had to decide about voting machines and made an interesting decision which de-facto banned them. They didn't really ban them, but simply set some requirements (apart from the usual as anonymity etc.) for voting machines which as of today none of them can meet:

"When electronic voting machines are deployed, it must be possible for the citizen to check the essential steps in the election act and in the ascertainment of the results reliably and without special expert knowledge."

So your average neighbor, grandma and co-worker must be able to understand how you came to the result because that's important in a democratic process. Your open-source voting machine software doesn't meet that requirement and neither does your fancy blockchain-based cryptographic system. But paper does: You make a cross on paper, you count them.

You can find an official, full English translation of the ruling here: http://www.bundesverfassungsgericht.de/SharedDocs/Entscheidu... (my quote is from paragraph 119).

Backstory: outsourcing our election administration to corporations. With the business model switching from bulk rates to per voter (whether they voted or not). That's the driver for all these disruptive changes.

With paper based precinct voting, perhaps 20% of the work is done outside. Mostly printing, ballots, pamphlets, poll books. But also buying some simple gear, like the optical (scanotronic style) scanners or punchcards (whatever is used).

Adding touchscreens to the mix outsourced more stuff.

With postal balloting, I'd estimate 50-60% is outsourced. For more printing. Signature matching. Ballot image scanners. A lot of bookkeeping software.

With internet voting, closer to 100% of the elections will be outsourced. What jurisdiction is sophisticated enough to build, run, maintain voting systems? None. So every jurisdiction will be like the state of Georgia, where all of election administration is conducted by a corporation.

--

Buddy of mine is trying to sell me on blockchain voting. Claims it'll protect privacy via some kinda tor like onion routing whatever.

I don't even care.

If we adopt some magical crypto voting system that actually work for our form of elections (unlikely), we'll completely forfeit the primary function of our governments, which is administrating elections.

Private voting and public counting will be gone.

> Private voting and public counting will be gone.

I may be missing your point or misunderstanding your comment, but what you describe there is a by-product of a successfully developed blockchain-based voting solution.

Everyone's votes are verifiably recorded, but publicly anonymous. You can still conduct public counts and everyone with access to the blockchain's ledger (or a trusted method of viewing the ledger) can confirm that the only votes that count towards the contest are those actually recorded. The additional benefit is that at any point in time, you can revisit your own vote to confirm it is still contributing to the outcome as it should.

I don't see how this in any way undermines or "forfeits the primary function of our government, which is administrating elections." This may be due to our differing opinions on what the role of government is. Generally my reasoning falls closer in line with Milton Friedman's (paraphrased) view:

> Government has three primary functions. It should provide for military defense of the nation. It should enforce contracts between individuals. It should protect citizens from crimes against themselves or their property.[...][1]

--- [1] http://www.goodreads.com/quotes/578626-government-has-three-...

> The additional benefit is that at any point in time, you can revisit your own vote to confirm it is still contributing to the outcome as it should.

Cryptography is not my strong point, so forgive me if this is obvious to others, but how can we build a system which simultaneously:

1. Allows you to verify that your vote counted in the direction you wanted it to

2. Allows you to be confident that your public "receipt" (half of a key, or whatever you use to confirm your vote counted) is unique, and you aren't sharing it with someone else who voted the same way (lumping multiple votes under the same "verifiable" vote)

3. Prevents you from publicly proving you voted in one direction, thus allowing you to sell your vote?

To put it another way, what good is it to be able to privately verify to myself that I voted in one direction if, when I find a discrepancy, I can't prove to people that the system miscounted my vote? Yet if I can prove it, what's stopping me from selling my vote?

As far as my research goes, we can make a secure system fairly easily which solves 1 and 2, but not 3. Getting all three may even be an impossibility theorem without physical control of the data (e.g. paper ballots). The cryptographic solutions are very new though, and there may be a breakthrough I haven't heard of yet. Still hopeful

"You can still conduct public counts..."

Do you have any examples of corporations submitting to FOIA (public record) requests?

My point is that our elections are being privatized. Internet voting is just the excuse.

"...my reasoning falls closer in line with Milton Friedman's..."

Mine's closer to Abraham Lincoln's:

  of the people, by the people, for the people

Elections are what give a democracy its legitimacy. The right to vote precedes all other rights. Like the Zeroth law of Robotics. It's unstated because it's assumed because we voted ourselves (nation) into existence.

> Do you have any examples of corporations submitting to FOIA..

No, I don't. "Public counts" in my comment refer to how the individual votes would be captured within a blockchain's ledger -- by design, visible and traceable.

I appreciate your follow-up. The right to free expression, and by association, the ability to vote, define our constitutional republic. It's so basic that I didn't consciously recognize it.

Is there anything that can solve this other than what you've described? They will grow and push indefinitely. It's like mass surveillance. Encryption works. It can just be impossible. The truly terrible outcome is if "internet voting" is both an excuse and becomes recognized as such. Then the public will fear "internet voting" for X decades. Even though it's ultimately the solution to this problem.

HUMINT is much more resource intensive than SIGINT. Paper voting is much more secure for persons than e-voting / polling.

Blockchains are more easily conducive to side-channel attacks at revealing who a person voted for than paper voting is.

I don't disagree you. I'm not aware of a production-grade blockchain voting protocol that yet exists that is resilient against one or more of the currently known de-anonymization attacks.

I'm not even sure it's possible to build one, given that humans are a portion of the workflow. On the other hand, I'd love to be proven wrong.

Any good suggestions about how blockchain systems propose to securely and anonymously distribute voting power?

In the paper system Michigan uses, ballots are not distributed anonymously, but the identifying information is torn off the ballot before it is counted and secured.

(I think it's pointless to propose a voting system that depends on individuals securely managing a cryptographic wallet; it's pretty clear we don't yet live in a world where that's a good idea. But it'd be interesting to see how they propose to solve the above problem.)

edit: logfromblammo answered my question below: https://news.ycombinator.com/item?id=12382298

How would blockchain voting address the ability to give your votes away prior to the election? (Or does it?)

For example, if I have one "votecoin" that I vote with that isn't tied back to me (as per necessity of private voting), how can the system ensure that it's actually me voting, and not just whoever I gave/sold my vote to beforehand?

Blockchain voting has yet to solve the biggest issue with voting: membership. How do we know that every vote came from a person who is supposed to be registered and that each person cast exactly one vote?

Hard problem.

> With electronic voting systems it is physically impossible to prevent manipulation and provide control.

Seems like a double standard. With paper it is impossible to prevent manipulation & provide control as well. To wit, the countless examples of election fraud:

https://en.wikipedia.org/wiki/Electoral_fraud

With either system is a question of risks & mitigation strategies. Neither system can guarantee the elimination of fraud.

It's about preventing easy scaling of voting fraud. To corrupt vote counting at multiple precincts when using a paper-based, manual counting system, requires the bribing of a lot of people in different locations, and is hard to keep quiet and go unnoticed the larger it gets. Using a digital voting system, if the software or firmware of any part of the voting process is compromisable, likely no-one will be able to tell, any many precincts affected.

I would guess that the writer started with a concept along to lines of "with electronic voting systems [that are connected to an external network], it is impossible to physically prevent manipulation and maintain control", which is true. I would guess that (s)he then allowed the hyperbolic urge to switched the formulation to "physically impossible." This is trivially false in a literal sense, an appeal to a qualitative notion of "impossible", which brings the argument into a context that is very familiar to Scotsmen and their imitators.

Within this context, there is indeed a double standard, as indeed it is no more possible to prevent a true Scotsman from committing electoral fraud with paper than to confound him with electrons. However, a claim may be misleading, unfair, or erroneous irrespective of whether it involves a double standard.

I believe that you make an error in seeking relief from the double standard. Essentially you are establishing that if we are to accept the first patently false assertion, we must equally accept the second (which is true)... and then agreeing to accept both.

So we are already quite far into the woods when we come to my reason for responding to you.

What you have presented is an example of the 'appeal to link' fallacy. You have suggested (and believed, probably) that there are countless examples of electoral fraud, and offered a link.

However, the document you have linked does not present or even suggest that there are countless examples of electoral fraud. It is a wikipedia article that enumerates specific methods of electoral fraud, loosely defined.

Furthermore, the document implicitly assumes the truth of the original assertion that is the subject of debate, namely that there are categories of fraud to which electronic voting systems are uniquely susceptible.

I was tempted to conclude that "this is why we can't have nice things", but the truth is that we could have much nicer things if we took a moment to step back and examine what we are discussing just a little more carefully.

Right. Even if the system is fraud proof. It must be so transparent so everybody understands how it works.

The US has not experienced (yet) a candidate that refuses to accept the results of an election. An unloyal candidate will claim fraud without any proof. And enough people would believe him to cause internal division or even a civil war or a coup d'etat.

It's really not possible to have a coup in the US. Government is too distributed for a quick take over. Civil war, on the other hand, could still happen, but I think it would probably end up with just a few stand-off situations and be resolved quickly.

I believe that the population of americans that view politics as reality TV are the majority.

Also, Bush v. Gore, florida election results, 2000. AKA "The Hanging Chad" incident. Also happens to be the year that americans at large learned the words "enjoin" and "disenfranchise."

I think you're assuming a coup against the government, which I'm not sure has ever happened anywhere. Any coup would come from a portion of the government, aligned with powerful interests external to the government.

As an example in the U.S., let's say that 3/4s of Republican governors, along with half of Republican Senators, and an absolute majority of the House, the vast majority of their individual financial backers, and a large number of generals decided that Trump lost due to voter fraud that occurred largely in states and cities with Democratic administrations. A group of them then read an open letter on Fox News about it, and the reading, while not carried on major networks, is broadcast on all channels by Comcast, Altice, and Cox Cable. As the letter is being read, a mass email goes out to the Trump email list to converge on the White House in DC and Clinton Foundation headquarters for a peaceful protest, named "Occupy the Crooked Election." As the thousands accumulate, the Governors that can announce that they're going to mobilize the state militias to do a recount, or rerun the election.

Is this impossible? How do you walk it back? Also, didn't we just have a massively blood soaked civil war 150 years ago?

For a succesful coup, you have to do a few things very quickly (hours),

  *Take over the media (simply not possible in America 
   due to the volume), initially to block 
   information 

  *Take over the Military (Too large and diverse 
   to accomplish in America.)

  *Arrest/Detain all political opposition.  
   (there are simply too many local governments 
   in America for this to happen.)

You have hours to accomplish all of this. If you can't manage it in hours, you will have an uprising and if you aren't quickly put down, you are in for a civil war.

Why would you even need #3? You simply make them persona-non-grata, or make them believe they are the bad guys (and hey, bad guys always complain about being bad guys!)

#1 is entirely feasible. #2 is entirely possible as well - there are lots of revolving doors with defense contractors and the DoD... heck, just read through the Patraeus scandal [1]. This a 4-star General (Commander of CentCom), Director of CIA and was a likely presidential candidate - just read through the tawdry and incestuous relationships within the community, and wonder at how easy it'd be to simply bribe everyone with an offer they couldn't resist.

[1] https://en.wikipedia.org/wiki/Petraeus_scandal

I posit this can be planned indefinitely. Once the hammer drops, yes countdown time is hours, but... take a look at the 2000 election results and the weeks it took to finally stop the clock on vote counting.

2 & 3 are why the coup in turkey failed.

Because the coup in turkey and foreign hacking of elections is so analogous...

There are only six[1] media corporations in the US.

[1]: https://en.wikipedia.org/wiki/Concentration_of_media_ownersh...

> A group of them then read an open letter on Fox News about it, and the reading, while not carried on major networks...

Fox isn't a major network? (Or did you mean "not carried on other major networks?)

State militia? Are you from the 1800s? The California State Army Reserve is 1500 strong. The California National Guard is 18000.

What's your point exactly? While the term "state militia" is a bit unusual these days, that's effectively what the National Guard is.

The National Guard doesn't report to California.

Is it really viable to apply old world paradigms to USA?

> It's really not possible to have a coup in the US

Then what exactly happened on 9/11? In non-banana variety republics, the person in charge of national air defense would have been minimally subjected to a court martial. This still happens here too [1] for far less serious events.

The general in this case, however, got promoted. [2]

> US. Government is too distributed for a quick take over

How about a very slow and deliberate take over. Would that work?

[1]: http://www.stripes.com/news/middle-east/another-navy-command...

[2]: http://www.jcs.mil/About/The-Joint-Staff/Chairman/General-Ri...

> How about a very slow and deliberate take over. Would that work?

No, that is not what a coup is.

That's just it, it isn't precise, at all. No one seems to remember 2000. It was a disaster in large part due to the lack of precision of paper voting. Those who don't remember are doomed to repeat I guess.

They were punchcards, not paper.

At this point, there is no reason not to use actual paper. Computer scanning and counting has gotten really good.

In Germany it's all paper and actual people doing the counting. I know Germany is not as big as the US, but it's not a small country. We usually have the first results at 6pm on voting day and the final results usually don't differ much from the first results.

That doesn't really help, though. With "marks on paper" voting, you don't have to deal with the hanging chad problem, but you get whole crops of people who don't fill the bubbles completely, or who partially mark two different bubbles, or a whole litany of other potential ambiguities.

I still agree that paper isn't vulnerable to the mass-scale undetectable vote-rigging that electronic machines have the potential to be! But it does have its own weaknesses: people who decided to switch to electronic definitely had credible reasons in mind at the time.

> With electronic voting systems it is physically impossible to prevent manipulation and provide control.

I don't know with a key tied to your vote that can be verified by yourself isn't half bad. Though then the exploit moves to the tallying of the votes since no one can independently do that...

I feel like electronic voting systems have to be the future but so many seem to think it's impossible it makes me wonder what the ultimate solution will end up being.

Isn't the blockchain technology that Bitcoin uses, perfect for this? It is 100% public, and you can review your vote easily.

Each person just needs a private key, and it could actually be issued once for each election.

However, I am not saying that we are ready for this, as not everyone is able to access the internet still.

That seems like a solution that could work. Granted someone could take your key and vote but they can already do that with paper ballets or a fake ID today anyway. I'm sure there are drawbacks but that's probably closer to the final solution.

We should have some scannable code on the voter registration cards that allow users to easily use their private key when they vote electronically.

all of it will be tallied up electronically regardless and you can manipulate it in any number of ways paper or not, challenged elections have shown as much.

so I don't understand this fantasy that "paper" which is an ambiguous term is bandied about as the ideal solution. From paper ballots to punch cards, there are hundreds of stories of fraud. If you don't just up and lose them or mysteriously get different votes than the district involved would imply. If you can't find fraud you simply challenge the votes until you get the results you want, keep some polls open than longer, or take people to vote and help them.

The ideal solution is getting it out of the hand of politicians into an independent board which can implement and secure it as needed.

The real reason this comes up is because one side or another wants an excuse for losing over and over at various levels. It also is because it is becoming even more difficult to hide fraud that they have to scream its more rampant than ever or lurking right around the corner.

Australia has already crossed the Rubicon with electronic counting of votes. The sweeping changes to the preference rules just before the last election also meant all the votes are tallied by Fuji Xerox in Canberra. They produced a video of the entire process and posted it to YouTube with locked comments. I have put a copy of this on Vimeo so others can comment on it as it is the scariest thing I have seen in a long long time. Please give it a watch even from 2:10 when the scanning in batches with the Windows PC is being done... https://vimeo.com/177354514

Seems like the best system is electronic that prints out a machine readable and human readable tape that the voter can verify, and in the event of a dispute, a human recount based on the tape is the final word on the result.

I think that the future should lie in digital democracy, where elections and referendums will be held constantly and this will require a well-working electronic system.

I have no faith in current crowd-validation elections.

Well then we will enter the era of constant populism. And politicians having permanently to validate everything against an every shifting public opinion. We will be rules by the press then, by clickbait headlines and populistic extremes I fear.

I do not believe, that people are able to vote for the longer term greater good if they are able to vote for everything in an instant.

Switz and Icelanders were able to pull it, did not they?

Of course this is a skill and it will be acquired for some time. After that, we don't even need politicians that much.

I'll try constant populism because today we live in the era of constant oppression.

Generally, populism implies oppression of the unpopular.

It's difficult enough (maybe impossible) to get a large chunk of the electorate to educate themselves about the issues once every 2, 4, or 6 years. You expect them to stay constantly up-to-date rather than choosing rational irrationality[1]?

[1] https://en.wikipedia.org/wiki/Rational_irrationality

No. I expect people to only vote on issues that they find pressing. By perhaps spending a replenishable "voting score".

That's just what we need. More aspects of our lives relegated to the political sphere, and more of our time devoted to motivating people to one side or the other of every issue.

But you also need to scale these systems. Paper doesn't scale

We did paper elections for ~100 million voters. How much more do you want to scale? Yes, it might take some time post elections to hash out results, but people don't take office the next day making that a non issue.

The benefit of transparent elections outweighs the costs of counting paper.

If we really wanted to do elections cheaply, we'd just hire a couple of undergrads to write a Node.js script and give polling stations some desktop PC's loaded up to the page. Would hardly cost anything. But doing elections cheaply is not the point.

For the USA, precinct-based paper mediated voting is the cheapest option.

We don't need PCs. Self contained scanotronic style systems (think SAT tests) work just fine. In fact, all but the largest jurisdictions (because of ballot size, complexity) could get away with manual counting.

Paper scales linearly. How it's done in the UK: http://www.bbc.co.uk/news/election-2015-england-32533064

That's cute. I'll post the same thing I post every time some non-US reader mentions how _their_ polity votes just fine with paper.

The last time I voted, I had a scantron ballot that was maybe A3 in size and needed filling both front and back. I'm probably forgetting a few offices, but I was voting for:

* President

* Federal Senator

* Federal House representative

* State Governor

* State Secretary of State

* State Senator

* State House representative

* Mayor

* City Council (slate)

* Sheriff

* State corporation commission (slate)

* Judges (slate)

* 5 or 6 ballot initiatives

* 1 or 2 local initiatives (i.e. school funding bond overrides or the like)

Are you highlighting a problem or verifying that a paper ballot worked fine?

18 US states currently do use paper ballots.

https://ballotpedia.org/Voting_equipment_by_state

Precincts are the smallest political jurisdiction. With poll sites, each precinct's election is administered independently. That's how paper scales.

How large is your precinct? Probably somewhere between 0 and 1000. Totally manageable.

And how large is your state? 1000 precincts? So you have to add up 1000 numbers (times number of candidates). Also totally manageable.

How frequently do you get a ballot that size? Are they all on four or two year terms?

Which is precisely why paper voting is advantageous. A compromised scalable system can have a much larger impact than a non scalable one.

There's an interesting analogy to things like WMDs here. Nukes can also have a much larger impact if abused or compromised, but we still built them. I expect eventually we'll have internet voting in place regardless if the benefits outweigh the potential downsides.

Population growth in the US is approaching 0.7% annually and has been declining for 2 decades. Not sure what scaling needs you foresee.

Fraud must NOT scale.

It's not like this stuff grows on trees, after all

Paper counting scales quite well. Infact it is trivially scalable.