12 ms·

Online Voting in U.S., Despite Risks, Should Be Developed

33 points by lambtron 14y ago

stretchwithme 14y ago
Just use multi-step authentication, one of which is sometime during the year showing up somewhere in person. Let one of the required authentication methods be designating a close friend or relative that is also a voter who can confirm that you actually voted. So they would get an electronic message with a url they would have to click after speaking with you and confirming that you did vote on the day in question. Make it so people have to be home to vote. And if we required the same process to confirm changes in address, we could cut down on identity fraud in general. Using multiple methods will make it harder to game the process. Its very hard to steal someone's password AND steal their phone AND kidnap a relative AND camp out at their house. Those that can't manage all of the required steps can keep voting the way they do now. They'll have much shorter lines to deal without everyone else there.
- tallanvor 14y ago
  I'm an expat... Showing up somewhere in person once a year is unlikely to be an option for me when you take into account that I visit the US every other year on average. And before you suggest using embassies as an option, you're assuming you live in a city with an embassy, or that you should be required to travel to one, regardless of the cost. How do you guarantee that the mail server will be up after you vote? Or that the server sending out the email won't be hit by a denial of service attack? For that matter, how do you determine if someone is at home? --Most people don't have fixed IP addresses, and they can be spoofed anyway. Finally, how do you determine that the vote logged in the database is what the person actually voted for and wasn't changed at a later time? With a paper ballot, you can always recount. Recounting database information is less easy.
thangalin 14y ago
A series of mock-ups showing one possibility for a policy-making system: http://imgur.com/a/PK69j http://imgur.com/a/PK69j
- 1qaz2wsx3edc 14y ago
  While I'm all for things like this, locality needs to be addressed. For instance, my local (city/state) issue, is maybe something the whole (country) should not vote on. Then again, baby steps right?
  - thangalin 14y ago
    Take a look at the URL in the mock-ups.
lindowe 14y ago
Have to say, I didn't really see any actual point being made by this author. Maybe its just a misleading headline, but as far as I could tell his argument boiled down to 1.) We can't let the Estonians beat us in the election technology race 2.) We live in the future, its time our election system was futuristic. The rest is just quotes from security experts about why he's wrong.
pdonis 14y ago
The solution to the one genuine problem mentioned in the article--that many people don't vote because the logistics are too difficult--is not online voting; it's early voting. (Which, btw, is what I believe President Obama was referring to when he said "we have to fix that" in reference to the long lines at the polls.) The article says the inherent security problems with online voting are "not impossible" to fix. In this sense, it's "not impossible" to keep Windows computers virus-free.
stcredzero 14y ago
As always, security is intimately tied to economics. How many people are interested in breaking into the Estonian elections, versus how many are interested in gaining power over the US elections? It's quite a stark contrast. This is why you can't lock your Ferrari in an rusty shed with a cheap master lock on it. (Though the shed will probably be excellent security for a 2nd hand bike.) E-voting in Estonia is probably a good idea. E-voting in the US is a catastrophically bad one until we get trusted execution infrastructure as described in Vernor Vinge's sci-fi books. (Yes, that's DRM, but DRM in the hands of individuals is very different from DRM in the hands of governments and large corporations.) EDIT: Right now, we don't need an ID to vote. That's because creating zombie voters isn't yet possible with paper technology, unless you game the counting, which has better security. Keeping such a conspiracy secret would be difficult, though. Gaming the system and getting away with it would be possible with e-voting machines. Given that only a few swing states need to be affected, organizations that can command 100's of millions of dollars could pull such a thing off.
- anewguy 14y ago
  It's funny that you mention this. In most American states, you don't even need ID to vote. We take vote security much less seriously than other countries.
- gojomo 14y ago
  FYI to sibling responder 'anewguy': looks like your comments are going auto-dead... so many readers can't see, reply, or vote on them. See the guidelines (http://ycombinator.com/newsguidelines.html http://ycombinator.com/newsguidelines.html) to inquire about this status.
- rmc 14y ago
  How many people are interested in breaking into the Estonian elections, versus how many are interested in gaining power over the US elections? You think there is no corruption or organised gangs or corrupt politicians in Estonia (or any other country)? Anywhere there is political office, someone'll want to scam their way in. Yes a small country can be less of a target, but they also need less votes to swing it, and have less money to protect against it.
specialist 14y ago
The punditry have learned nothing. Help America Vote Act (HAVA) ushered in the touchscreens in a massive turf grab. Insecure, unreliable, untested touch screens were rushed to market. The same forces are now at work pushing vote by mail and then electronic balloting. For the vendors, is about charging for the upgrade and switchover costs. Anything to make a buck. Election integrity is barely considered. During the hearings to permit casting ballots VIA EMAIL in my state, a retired general now lobbying for the vendors claimed "internet security has gotten really good." Well, shit, if a general says it's cool, that's good enough for me. Bill sailed through legislature unanimously. Because its for the children, err, troops. I talked to many of the legislators about electronic voting and casting ballots via the internet. First, they didn't understand that email was delivered via the internet in the clear. Second, if you're explaining, you're losing. Meaning once you start talking policy, their eyes glaze over, oops, sorry, your five minutes are up. Next! The only way the legislators vote against this stuff is if they feel the heat. Meaning stuff the hearings and their inboxes with opposition. I could go on and on, but I have pancakes waiting for me.
- dmckeon 14y ago
  On-line voting is not impossible, but people tend to see the "I marked my ballot" part and ignore the rest of the voting ecosystem. Registering voters implies maintaining a registration DB, and handling moves, deaths, changes of name, changes of civil status that may disenfranchise or re-enfranchise a voter. Preparing ballots, paper or otherwise, implies matching voters to the appropriate ballot, precinct, or voting center. Avoiding or invalidating multi-voting (in the same jurisdiction) implies knowing whether a voter has or has not voted, and keeping them from voting multiple times. Counting paper ballots implies accounting for spoiled and unused ballots and dealing with faint or unintended marks. Secrecy of ballot implies that nobody else knows or can find out how a voter actually voted (currently unless they show someone a mail-in ballot before mailing it.) Imagine the result of a smartphone app that wrapped around a voting app, and offered to pay voters $50 to vote for or against a candidate or ballot measure - even if the payoff app completely failed to work, just the purported existence of such a tool might cast doubt on a election result. Voting online is not a "greenfield" problem - it would have to subsume a lot of previous experience and knowledge before being thoroughly trusted by its users. I would suggest to anyone exploring online voting that they get a grassroots look at the current process - volunteer to be a elections clerk (judge, watcher, etc.) at your next local election, and/or become a voter registrar if your jurisdiction allows citizens to fill that role. If you'd rather just web-browse, I can recommend https://en.wikipedia.org/wiki/Avi_Rubin https://en.wikipedia.org/wiki/Avi_Rubin as a useful and thoughtful source. PS - my voting preference is that voters mark scannable ballots, that are scanned immediately in the voter's presence. You get the persistence of paper, immediate feedback for mismarks or unscannable ballots, and good reproducibility of count if needed. Also, if a surge of people arrives at a polling place, paper ballot marking can be done in parallel, rather than having a bottleneck of a few (expensive) machines that all voters must use (registration look-up and ballot scanning are still bottlenecks).
KaoruAoiShiho 14y ago
There is 1 good reason for electronic voting (among many reasons against). If electronic voting is implemented then conceivably it would be much easier to convince your base to actually cast their ballots, reducing emphasis on the "ground game" and having a big operation, and thus reducing costs, and more easily allowing third parties to be viable and less well funded candidates to make a bigger impact during primaries.
naturalethic 14y ago
This already exists. It's call shopping.
jellicle 14y ago
Voting online would be trivial. I use "trivial" in the hacker sense to mean a problem that, while it might be difficult, large and lengthy, has already been thoroughly explored and a set of solutions and best practices is known, and should easily be accomplished by implementors of average or above skill levels. We do things similar to online voting all the time. When Superbowl or Lady Gaga tickets go on sale, many people flood limited database servers and have to be put in a queue. I can transfer hundreds of thousands of dollars with my smartphone - that's not a problem, but a vote which is worth nothing is going to be a target? When Census forms are sent out, a code is sent by physical mail to addresses - that code, when entered online, can be used to fill out the form. Or just fill out the paper form and drop it in the mail. Trivial either way. For that matter, thousands of organizations, not just Estonia, have binding internet votes every year. When you read about people waiting seven hours in line, it's not unforeseen problems or incompetence; it's malice. If you show me a story about people waiting seven hours in line, I'll show you a Democratic district in a state with Republican, partisan election officials. Every time. Those sorts of problems can't be solved by technology; they're people problems, not tech problems. Right now, internet voting is a people problem, not a tech problem. Internet voting might encourage more people to vote - and that's why Republicans are against it.
- baddox 14y ago
  > When you read about people waiting seven hours in line, it's not unforeseen problems or incompetence; it's malice. If you show me a story about people waiting seven hours in line, I'll show you a Democratic district in a state with Republican, partisan election officials. I'd love a citation on this, or at least a few anecdotal examples. Granted, it does fit with the general perception of party strategies (Democrats try to buy voters, Republicans try to prevent voters), but I'd still like to see some evidence.
- cabalamat 14y ago
  > Voting online would be trivial. I disagree. Online voting is highly non-trivial once you factor in that any voting system must be (1) verifiably accurate, and that (2) voting must be secret.
mickgardner 14y ago
Paper and pencil based voting, despite the risks, should be developed in USA, shouldn't take to long.
mindslight 14y ago
Hey tech community! Remember the time before the eye of Sauron gazed here, demanding a superficial veneer of progress to demonstrate "its" society's advancement? Remember when one informed and well-reasoned opinion was worth more than twenty idiotic and entitled ones? (kind of the opposite of democracy. oops, that wasn't meant to last!) Remember when we dreamed of engineering actual solutions, rather than shoehorning all problems into what can be solved using html+http+database, resultant deficiencies be damned? Well I guess only about a tenth of you do, and of those only a tenth care. sigh.
mrgordon 14y ago
I'll just point out that top security experts like Ronald Rivest (co-inventor of RSA) expressed extreme skepticism that we would be able to implement electronic voting without seriously endangering our elections during the MIT-Caltech voting project. I'm not trying to discourage anyone from working towards it, but it has a lot of issues.
onerealkewlguy 14y ago
I believe that the only way that online or electronic voting could work would be as an freely distributed "open source" community built & tested solution. If open source is secure enough to run the Whitehouse website it is likely secure enough to manage recording a vote. Maybe we could engineer a situation where we use thin virtual clients from a master server and the actual "complete operating system" of the voting device could be imaged or have a current snapshot stored on the same media as the vote reciept itself for possible later verification or even independent verification on the spot by interested code developers.
anonymouz 14y ago
The article is just mind-bogglingly silly: The author quotes a number of experts, explaining why it is a really bad idea and why it would compromise the integrity of the election. Then he goes and, around those quotes, puts some fluff where he suggests we should do it anyway, cause he'd find it convenient to vote on his smartphone. Here's a basic check for the author of the article: If you want to demonstrate how online voting is the future, but you can't find any experts supporting your point of view, and don't have anything refuting their serious arguments against it, maybe it just isn't such a good idea after all.
- 001sky 14y ago
  The issue is the fraud cannot be undone. Unlike your IRS taxes online, there is no IRS to audit trail the vote. And even if there were (cost/privacy logistics aside), the harm is irreprable. If the wrong guy is given power for 2-3 years, its too late to then switch over (justice delayed=justice denied). That is the nutshel of why this "obvious" idea is fatally flawed, IMHO. You are concentrating political power in a black-box not subject to checks/balances, etc. Odd, but paper ballots are the simpler/smarter choice. We just need to lose the clowns administering the elections. Perhaps this should be handed off to a the judiciary or something? Dunno. A good Compromise seems to be (1) national holiday to vote (to increase turnout); (2) IDs (to maintain semblance of legitimacy); (3) simple paper ballot (secure, private, verifiable); (4) computerized registration (to ensure any problems sorted out in advance re(#2).

33 points by lambtron 14y ago

Just use multi-step authentication, one of which is sometime during the year showing up somewhere in person. Let one of the required authentication methods be designating a close friend or relative that is also a voter who can confirm that you actually voted. So they would get an electronic message with a url they would have to click after speaking with you and confirming that you did vote on the day in question. Make it so people have to be home to vote. And if we required the same process to confirm changes in address, we could cut down on identity fraud in general. Using multiple methods will make it harder to game the process. Its very hard to steal someone's password AND steal their phone AND kidnap a relative AND camp out at their house. Those that can't manage all of the required steps can keep voting the way they do now. They'll have much shorter lines to deal without everyone else there.

tallanvor 14y ago
I'm an expat... Showing up somewhere in person once a year is unlikely to be an option for me when you take into account that I visit the US every other year on average. And before you suggest using embassies as an option, you're assuming you live in a city with an embassy, or that you should be required to travel to one, regardless of the cost. How do you guarantee that the mail server will be up after you vote? Or that the server sending out the email won't be hit by a denial of service attack? For that matter, how do you determine if someone is at home? --Most people don't have fixed IP addresses, and they can be spoofed anyway. Finally, how do you determine that the vote logged in the database is what the person actually voted for and wasn't changed at a later time? With a paper ballot, you can always recount. Recounting database information is less easy.

thangalin 14y ago

A series of mock-ups showing one possibility for a policy-making system: http://imgur.com/a/PK69j http://imgur.com/a/PK69j

1qaz2wsx3edc 14y ago
While I'm all for things like this, locality needs to be addressed. For instance, my local (city/state) issue, is maybe something the whole (country) should not vote on. Then again, baby steps right?
- thangalin 14y ago
  Take a look at the URL in the mock-ups.

lindowe 14y ago

Have to say, I didn't really see any actual point being made by this author. Maybe its just a misleading headline, but as far as I could tell his argument boiled down to 1.) We can't let the Estonians beat us in the election technology race 2.) We live in the future, its time our election system was futuristic. The rest is just quotes from security experts about why he's wrong.

pdonis 14y ago

The solution to the one genuine problem mentioned in the article--that many people don't vote because the logistics are too difficult--is not online voting; it's early voting. (Which, btw, is what I believe President Obama was referring to when he said "we have to fix that" in reference to the long lines at the polls.) The article says the inherent security problems with online voting are "not impossible" to fix. In this sense, it's "not impossible" to keep Windows computers virus-free.

stcredzero 14y ago

As always, security is intimately tied to economics. How many people are interested in breaking into the Estonian elections, versus how many are interested in gaining power over the US elections? It's quite a stark contrast. This is why you can't lock your Ferrari in an rusty shed with a cheap master lock on it. (Though the shed will probably be excellent security for a 2nd hand bike.) E-voting in Estonia is probably a good idea. E-voting in the US is a catastrophically bad one until we get trusted execution infrastructure as described in Vernor Vinge's sci-fi books. (Yes, that's DRM, but DRM in the hands of individuals is very different from DRM in the hands of governments and large corporations.) EDIT: Right now, we don't need an ID to vote. That's because creating zombie voters isn't yet possible with paper technology, unless you game the counting, which has better security. Keeping such a conspiracy secret would be difficult, though. Gaming the system and getting away with it would be possible with e-voting machines. Given that only a few swing states need to be affected, organizations that can command 100's of millions of dollars could pull such a thing off.

anewguy 14y ago
It's funny that you mention this. In most American states, you don't even need ID to vote. We take vote security much less seriously than other countries.
gojomo 14y ago
FYI to sibling responder 'anewguy': looks like your comments are going auto-dead... so many readers can't see, reply, or vote on them. See the guidelines (http://ycombinator.com/newsguidelines.html http://ycombinator.com/newsguidelines.html) to inquire about this status.
rmc 14y ago
How many people are interested in breaking into the Estonian elections, versus how many are interested in gaining power over the US elections? You think there is no corruption or organised gangs or corrupt politicians in Estonia (or any other country)? Anywhere there is political office, someone'll want to scam their way in. Yes a small country can be less of a target, but they also need less votes to swing it, and have less money to protect against it.

specialist 14y ago

The punditry have learned nothing. Help America Vote Act (HAVA) ushered in the touchscreens in a massive turf grab. Insecure, unreliable, untested touch screens were rushed to market. The same forces are now at work pushing vote by mail and then electronic balloting. For the vendors, is about charging for the upgrade and switchover costs. Anything to make a buck. Election integrity is barely considered. During the hearings to permit casting ballots VIA EMAIL in my state, a retired general now lobbying for the vendors claimed "internet security has gotten really good." Well, shit, if a general says it's cool, that's good enough for me. Bill sailed through legislature unanimously. Because its for the children, err, troops. I talked to many of the legislators about electronic voting and casting ballots via the internet. First, they didn't understand that email was delivered via the internet in the clear. Second, if you're explaining, you're losing. Meaning once you start talking policy, their eyes glaze over, oops, sorry, your five minutes are up. Next! The only way the legislators vote against this stuff is if they feel the heat. Meaning stuff the hearings and their inboxes with opposition. I could go on and on, but I have pancakes waiting for me.

dmckeon 14y ago
On-line voting is not impossible, but people tend to see the "I marked my ballot" part and ignore the rest of the voting ecosystem. Registering voters implies maintaining a registration DB, and handling moves, deaths, changes of name, changes of civil status that may disenfranchise or re-enfranchise a voter. Preparing ballots, paper or otherwise, implies matching voters to the appropriate ballot, precinct, or voting center. Avoiding or invalidating multi-voting (in the same jurisdiction) implies knowing whether a voter has or has not voted, and keeping them from voting multiple times. Counting paper ballots implies accounting for spoiled and unused ballots and dealing with faint or unintended marks. Secrecy of ballot implies that nobody else knows or can find out how a voter actually voted (currently unless they show someone a mail-in ballot before mailing it.) Imagine the result of a smartphone app that wrapped around a voting app, and offered to pay voters $50 to vote for or against a candidate or ballot measure - even if the payoff app completely failed to work, just the purported existence of such a tool might cast doubt on a election result. Voting online is not a "greenfield" problem - it would have to subsume a lot of previous experience and knowledge before being thoroughly trusted by its users. I would suggest to anyone exploring online voting that they get a grassroots look at the current process - volunteer to be a elections clerk (judge, watcher, etc.) at your next local election, and/or become a voter registrar if your jurisdiction allows citizens to fill that role. If you'd rather just web-browse, I can recommend https://en.wikipedia.org/wiki/Avi_Rubin https://en.wikipedia.org/wiki/Avi_Rubin as a useful and thoughtful source. PS - my voting preference is that voters mark scannable ballots, that are scanned immediately in the voter's presence. You get the persistence of paper, immediate feedback for mismarks or unscannable ballots, and good reproducibility of count if needed. Also, if a surge of people arrives at a polling place, paper ballot marking can be done in parallel, rather than having a bottleneck of a few (expensive) machines that all voters must use (registration look-up and ballot scanning are still bottlenecks).

KaoruAoiShiho 14y ago

There is 1 good reason for electronic voting (among many reasons against). If electronic voting is implemented then conceivably it would be much easier to convince your base to actually cast their ballots, reducing emphasis on the "ground game" and having a big operation, and thus reducing costs, and more easily allowing third parties to be viable and less well funded candidates to make a bigger impact during primaries.

naturalethic 14y ago

This already exists. It's call shopping.

jellicle 14y ago

Voting online would be trivial. I use "trivial" in the hacker sense to mean a problem that, while it might be difficult, large and lengthy, has already been thoroughly explored and a set of solutions and best practices is known, and should easily be accomplished by implementors of average or above skill levels. We do things similar to online voting all the time. When Superbowl or Lady Gaga tickets go on sale, many people flood limited database servers and have to be put in a queue. I can transfer hundreds of thousands of dollars with my smartphone - that's not a problem, but a vote which is worth nothing is going to be a target? When Census forms are sent out, a code is sent by physical mail to addresses - that code, when entered online, can be used to fill out the form. Or just fill out the paper form and drop it in the mail. Trivial either way. For that matter, thousands of organizations, not just Estonia, have binding internet votes every year. When you read about people waiting seven hours in line, it's not unforeseen problems or incompetence; it's malice. If you show me a story about people waiting seven hours in line, I'll show you a Democratic district in a state with Republican, partisan election officials. Every time. Those sorts of problems can't be solved by technology; they're people problems, not tech problems. Right now, internet voting is a people problem, not a tech problem. Internet voting might encourage more people to vote - and that's why Republicans are against it.

baddox 14y ago
> When you read about people waiting seven hours in line, it's not unforeseen problems or incompetence; it's malice. If you show me a story about people waiting seven hours in line, I'll show you a Democratic district in a state with Republican, partisan election officials. I'd love a citation on this, or at least a few anecdotal examples. Granted, it does fit with the general perception of party strategies (Democrats try to buy voters, Republicans try to prevent voters), but I'd still like to see some evidence.
cabalamat 14y ago
> Voting online would be trivial. I disagree. Online voting is highly non-trivial once you factor in that any voting system must be (1) verifiably accurate, and that (2) voting must be secret.

mickgardner 14y ago

Paper and pencil based voting, despite the risks, should be developed in USA, shouldn't take to long.

mindslight 14y ago

Hey tech community! Remember the time before the eye of Sauron gazed here, demanding a superficial veneer of progress to demonstrate "its" society's advancement? Remember when one informed and well-reasoned opinion was worth more than twenty idiotic and entitled ones? (kind of the opposite of democracy. oops, that wasn't meant to last!) Remember when we dreamed of engineering actual solutions, rather than shoehorning all problems into what can be solved using html+http+database, resultant deficiencies be damned? Well I guess only about a tenth of you do, and of those only a tenth care. sigh.

mrgordon 14y ago

I'll just point out that top security experts like Ronald Rivest (co-inventor of RSA) expressed extreme skepticism that we would be able to implement electronic voting without seriously endangering our elections during the MIT-Caltech voting project. I'm not trying to discourage anyone from working towards it, but it has a lot of issues.

onerealkewlguy 14y ago

I believe that the only way that online or electronic voting could work would be as an freely distributed "open source" community built & tested solution. If open source is secure enough to run the Whitehouse website it is likely secure enough to manage recording a vote. Maybe we could engineer a situation where we use thin virtual clients from a master server and the actual "complete operating system" of the voting device could be imaged or have a current snapshot stored on the same media as the vote reciept itself for possible later verification or even independent verification on the spot by interested code developers.

anonymouz 14y ago

The article is just mind-bogglingly silly: The author quotes a number of experts, explaining why it is a really bad idea and why it would compromise the integrity of the election. Then he goes and, around those quotes, puts some fluff where he suggests we should do it anyway, cause he'd find it convenient to vote on his smartphone. Here's a basic check for the author of the article: If you want to demonstrate how online voting is the future, but you can't find any experts supporting your point of view, and don't have anything refuting their serious arguments against it, maybe it just isn't such a good idea after all.

001sky 14y ago
The issue is the fraud cannot be undone. Unlike your IRS taxes online, there is no IRS to audit trail the vote. And even if there were (cost/privacy logistics aside), the harm is irreprable. If the wrong guy is given power for 2-3 years, its too late to then switch over (justice delayed=justice denied). That is the nutshel of why this "obvious" idea is fatally flawed, IMHO. You are concentrating political power in a black-box not subject to checks/balances, etc. Odd, but paper ballots are the simpler/smarter choice. We just need to lose the clowns administering the elections. Perhaps this should be handed off to a the judiciary or something? Dunno. A good Compromise seems to be (1) national holiday to vote (to increase turnout); (2) IDs (to maintain semblance of legitimacy); (3) simple paper ballot (secure, private, verifiable); (4) computerized registration (to ensure any problems sorted out in advance re(#2).